The following are some of the frequently asked questions about the Tandem Computers – HP NonStop servers – that I have gathered from the Google Group comp.sys.tandem and the Yahoo Group Tandem_Computers – two of the most useful usenets currently available for discussions on Tandem Computers!
Note that all this information has been compiled from publicly available sources, and no violation of copyrights/IPRs is intended. If you find any of your stuff which should not be here, please let me know.
Additionally, if you can answer any of the unanswered questions, or would like to improve any of the existing answers, you are more than welcome to contact me. All due credit will be given.
The best way to search your way through this is Ctrl + F!
This is a dump from 2008. USE AT YOUR OWN RISK! I am looking for somebody who can keep the below FAQ current. Ideally somebody who is actively working on the Tandem environment currently. Let me know using the comments and we can discuss.
– What are Tandem computers?
Tandem computers were established in 1974 by James Treybig (“Jimmy T.”) and became a part of Compaq computers corp. in June 1997. Compaq’s NonStop Himalaya S-series servers provide 24 X 7 uptime that global infrastructure industries expect. Tandem Computers are for those for whom the difference between 99.9 and 99.999% availability is a BIG difference.
View the merger story at http://news.cnet.com/news/0-1003-200-321758.html
– What is its USP?
The USP of Tandem computers is their very high uptime, very high availability and reliability that is typically required in securities transactions, credit-card and ATM transactions, cellular phone services, etc.
– What is the Tandem file system?
Tandem, like many Linux and Unix flavors, treats disk files, devices like terminals, printers, tape drives, and processes, as files. Disk files can be either SQL files or Enscribe files. You can use NonStop SQL to access the former type and Enscribe database record manager to access the later type of files. Enscribe files can further be key-sequenced, entry-sequenced, relative or unstructured files. The file system allows you to open devices and processes as files and do the communication.
Besides, the usable space on the disk is called volume and the volume can be assumed to be divided into subvolumes, which in turn store the files. There is no concept of subvolume within a subvolume.
– Is a web-server available for Tandem?
iTP WebServer from Tandem is a web-server for Tandem. Besides, a very simple server can be downloaded fromhttp://126.96.36.199:5555/kku/tools.html
– What Tandem 6530 emulators are there?
Many (especially for Windows). See http://www.madoreconsulting.com/html/tandemlinks-term.htm.
For DOS: PC6530 by Tandem.
For Windows: Outside View by Crystal Point.
Win6530 CAIL …
For Macintosh: MacUWS?
For X-Windows: X6530 by Tandem.
NSK6530 from Bowden Systems (http://www.bsi.com/)
Java based 6530 emulators MR-Web6530 and J6530 from ComForte (http://www.comforte.com/)
Do Tandem computers have a familiar user interface?
Yeah! The OSS(Open System Services) face of Guardian is very much Unix-like.
– How do Tandem computers compare with other popular platforms with respect to up-time and reliability?
??View the pic
– What are the networking protocols supported by Tandem computers?
Tandem applications can involve a vairety of terminal types, printers, workstations, and personal computers, and devices such as ATMs, cash registers, robots and bar-code readers. These devices can be connected to Himalaya S-series servers over a number of media: dial-up lines, leased lines, digital networks, LANs, and frame relay and S.25 PSDNs.
Tandem offers products that implement the protocols of OSI(open systems interconnection), TCP/IP, NetBIOS. Besides Tandem products also allow support for IBM SNA(Systems Network Architecture) applications and devices.
– Can a Tandem network be spread over a large geographical area?
Yeah! You can use EXPAND or FOX(Fibre Optic eXtension) Tandem products for the purpose.
– What are the processors that Tandem computers run on?
Tandem computers run on MIPS processors. Visit MIPS Publications page.
– Is learning programming for Tandem a start-from-scratch experience?
Nope! C, C++, Java and COBOL85 programming is possible on Tandem. These languages have been enhanced to implement support for the Tandem fault-tolerance principles.
Besides, proprietary languages like TAL(Transaction Application Language) and screen-based Pathway application programming language SCOBOL(Screen COBOL) are also available but owning to their similarity to C and COBOL respectively, are not too tough to learn.
– What database support does Tandem provide?
Tandem has ENSCRIBE and SQL databases.
– Can you run multiple applications on a Tandem machine?
This surprising question comes up once in a while. Rest assured. You can!
– Is there a file compression program for Tandem?
OSS contains the standard UNIX compress.
Guardian Compress from Bowden Systems: http://www.bsi2.com/download-free.html
TZip from Proficient Software: http://prof-software.hypermart.net/
ZIP for Tandem: ftp://sunsite.cnlab-switch.ch/mirror/infozip/TANDEM/
PAK/UNPAK is available from http://188.8.131.52:5555/kku/tools.html
COMPRESS for Guardian is available from Computer Security Products Inc. http://www.compsec.com/
– How to give access to Tandem applications from the web?
The Tandem iTP Webserver product and other products are available from Compaq. In fact, many of the applications already have web interface like Web-SCF, Web-DSM/SCM, etc. The iTP Webserver is available for free download at http://oss.atc-compaq.com/
– How to set time from an external clock?
There are some sample programs that use TCP/IP at http://184.108.40.206:5555/kku/gettime.html
– How to synchronize time between Tandem systems?
Either set the time of all machines from an external clock or use a program like NSX (Network Statistics Systems Extended, T9683, included with NSK).
– Are there Tandem resources on the web?
Well, what are you looking at?
There are some serious Tandem links at http://www.madoreconsulting.com/html/tandemlinks.htm You may also see the left browsing pane of this page to go to some Tandem links.
– Are there any books on Tandem?
Not one that I know of! Tandem and Pathway are mentioned in many books on transaction processing, fault tolerant computing and database systems. For all practical purposes the Total Information Manager (TIM) CD available from Compaq is the best source of all the relevant information about Tandem computers and also the applications that run on them.
– Is there a way to send email (SMTP) from Tandem?
Yeah! An application that does this magic, with its source in C, is available under the “C Programs” link in the left browsing pane on this page. The How-To document for the same is available here.
-How to I display BIG BIG letters on the terminal using TACL/in the OSS environment?
In Guardian environment the following TACL macro will do the magic for you. Just type “RUN BIGLETER HELLO” on the TACL prompt and HELLO will be printed on the screen in the BIG BIG letters. In the OSS environment the same can be accomplished by “osh -p “/bin/banner” Hello” command which displays Hello in the big letters on the terminal.
How do you MOVE files from one subvolume to another on a Tandem system? Is there a FUP MOVE command? How do you rename a subvolume on a Tandem system?
(A) Suppose you want to MOVE “all” files from a subvolume $DATA00.FIRST to the subvolume
FUP RENAME $DATA00.FIRST.* , $DATA00.SECOND.*
You may replace the wild-card * with any file-name template. This command is as good as renaming the subvolume FIRST to SECOND.
How do you extract preliminary information from a SAVEABEND file?
(A) On Inspect prompt issue the command
–INFO SAVEFILE FILE ZZSA3126
What is Garth?
(A) Garth is a symbolic, high-level tool that does low-level debugging, working on both live system (via TCP-IP) as well as processor images (dumps). It includes features of gdb, Crunch, as well as powerful, new utilities. Garth was originally the brainchild of Hieu Tran when he was working on Native Mode 1. Based upon the GNU Debugger (GDB) as well as the Tcl/Tk scripting language, it is easily flexible and extensible, as users can write their own scripts and invoke them when desired. Garth is available for various UNIX platforms including Sun-OS, Solaris, as well as for Tandem NSK. Details available on Garth home-page at http://nskernel.web.tandem.com/garth/
What is HIGHPIN?
(A) Each process that runs in a CPU has a Process Indentification Number (PIN) that uniquely identifies it. Before the Dxx release, there was a limit of 256 processes that could run at the same time in a CPU. A PIN could take a value between 0 and 255. The Dxx release increased the limit on the number of processes that can run in a CPU. (It also increased many other limits). Processes that run in PINs greater than 255 are called HIGHPIN processes, the others are called LOWPIN processes – what else 🙂 This is all fine and dandy, except that programs may require changes in order to run in a high pin. This is because HIGHPINs are stored as 16 bit integers, and LOWPINS are stored as 8 bit integers. The Guardian procedure calls that are used for process management were revamped. The old calls are still there, but for the most part they can only be used on LOWPIN processes. If a program wants to run in a HIGHPIN, or wants to communicate with another process that runs in a HIGHPIN, it will need to be changed to use the new procedure calls. Also, there are a new set of system messages (that are read from $RECEIVE) that relate to HIGHPIN processes. Again, the old messages are still there, but are only useful for LOWPIN processes.
What are the differences in the pointers in Tandem C and TAL?
(A) Differences between TAL and C pointers include the following:
·TAL structure pointers can point to a byte or word address.
· C structure pointers always point to a word address. To pass a C structure pointer to a TAL routine that expects a byte structure pointer, you must explicitly
cast the C pointer to type char.
· TAL pointers are dereferenced implicitly.
· C pointers are usually dereferenced explicitly.
· Small-memory-model C routines use 16-bit pointers only.
· Large-memory-model C routines use 32-bit pointers only, even if the pointers refer to the user data segment. In global structure declarations, you must specify _lowmem in the storage class of the declaration.
· If a TAL routine expects a 16-bit pointer, the C pointer you pass must refer to an object in user data space.
What is BASE24? I see it everywhere from job postings to Tandem developers’ resumes;-)
Many customers around the world use BASE24 to manage devices, route and switch transactions, and provide authorization support for high-volume payments processing. The software operates on Compaq NonStop Himalaya systems to provide 24/7 support for ATM and POS networks, manned teller systems, telephone banking, mobile commerce, and Internet banking and commerce. ACI (www.aciworldwide.com) develops and markets many BASE24 based applications like BASE24-ATM, BASE24-CARD, BASE24-BILLPAY, BASE24-TELLER, etc.
How do you determine the RAM on a Tandem machine?
Type the following at the tacl prompt :
> peek /cpu 0/ paging
Here’s a sample output :
PEEK – T9050G09 – (04APR01) SYSTEM \WHIT10
COPYRIGHT TANDEM COMPUTERS INCORPORATED 1981, 1985, 1988, 1990
22 SEP 2001, 18:20___ELAPSD 32:11:07___CPU 0(NSR-W)
PAGES: PHYSCL SWAPBL FREE FREEMIN FREEQUOTA FREERED LOCKED LOCKED(KSEG0)
(16Kb) 16384 16164 867 10 10 5 7372/14144 7372/28480
FAULTS ALLOCS DISKREADS DISKWRITES MUTEXCRAX NONMUTEXCRAX
TOTAL 90718 101394 46526 1810 21017 649046
(per sec) 0.78 0.87 0.40 0.01 0.18 5.60
REDHIT REDBUSY REDTASK
TOTAL 0 0 0
(per sec) 0.00 0.00 0.00
CLEANQ: FULLS FRLST:HITS CLOCK:CALLS FAILS CYCLES ALIASES: FAILS
0 0 100202 2745 0 3.58 0 0
This will give you the amount of RAM for CPU 0 :
PAGES * PHYSCL = (16kb) * 16384= 256 MB (rounded)
How do I know the full Guardian filename of a file under OSS?
Use “GNAME <filename>” on the OSS prompt.
How do you run TACL commands from OSS and vice-versa?
As the kernel is same for both OSS and Guardian, there is an option of executing TACL commands on the OSS prompt and the other way round.
This is the syntax of running TACL commands on OSS :
gtacl -c ” < tacl cmd > ”
E.g. gtacl -c “fup copy taclcstm”
Here’s the way of running OSS commands on TACL prompt :
OSH -c ” < OSS CMD > ”
E.g. If you want to see the contents of a file, you can try this :
osh -c “more taclcstm”
What are BIG files?
On Tandem, Enscribe and SQL are the (only) two ways by which an application program can maintain its data.
Enscribe, with the help of DP2 disk process, manages database of an application in the form of records. It also maintains data integrity in case of fatal errors.
Initially, Enscribe was designed to support files upto 2GB size. Since disk drives available at that time were of few 100 MBs only, a limit of 2GB was found sufficient. Moreover, since integers were used to hold record pointers and EOF information, 2GB could fit well into a signed integer (2^15 = 2GB).
To suffice application needs of huge data files (files larger than 2GB), concept of partitioning was introduced. But there is a maximum limit of 16 partitions for a file in Enscribe. SQL, which was introduced later, had no such limits. Moreover, SQL allows adding and deleting partitions to Key Sequenced tables, which is not possible with Enscribe Key Sequenced Files.
With the invention of disk drives greater than 2GB (Disks of size 72GB disks are expected in future), 2GB limit seemed to be a restriction on Enscribe, which needed to be overcome. With D46 release of NSK for K-Series Systems (and G06 for S-Series Systems), Enscribe supports unpartitioned files of size 1TB (1024 GB).
The new (Big File) version of file is called Fomat 2 file, whereas the older version has been termed as Format 1.
Note: Format 2 File Concept is valid only for database files. EDIT Files (Code 101), Object Files (Code 100 or 700) and all other files having reserved Code Number (Code Numbers Between 100 and 999) cannot be Format 2.
What is the meaning of Primary Extents, Secondary Extents and MAXEXTENTS?
An extent is a set of n disk pages (n = value of extent). An enscribe file is created with extents.
A disk page represents 2K of data (= 2048 bytes).
Extents are allocated in consistent data blocks on the physical disk drive.
Primary extent: The disk space ‘booked’ when the first data is written to the file. The size is defined by the value of Primary Extent Size.
Secondary extent: Whenever the first/previous extent is filled up, new disk space is ‘booked’. The size is defined by the value of Secondary Extent Size.
MAXEXTENTS: The maximum value of extents to allocate for a file (1 primary + 1 or more secondary extents).
The Primary extent is the amount of physical disk space allocated for a file when the first piece of data is written to it. When that space is filled, a Secondary extent is allocated based on the size given in the file description. The maximum number of extents for a file is determined by the value of MAXEXTENTS, and is the total number that will be allocated to a file before it is considered “full”.
Example: Creating a file with Ext (5000, 3000), MaxExtents 16 means:
The first write will allocate (reserve or book if you want) 5000 pages on the physical disk = 5000 * 2048 = 10.240.000 bytes (= ~10MB). Once you filled the ~10MB of data, a next extent is allocated. The size is 3000 pages on the physical disk = 3000 * 2048 = 6144000 (= ~6MB). Same scenario when this extent is filled. MAXEXTENTS delimits the maximum file size. In this example the maximum data size all in all will be (5000 + (15 * 3000)) * 2048 = 50000 * 2048 = 102400000 (= ~100MB)
Notes: Allocating extents takes time. The bigger the data income/growth, the bigger the extents should be. If you know you have a lot of data use a big primary extent so that the seek time is reduced because the data will be stored in a physically contiguous space. A once allocated extent is not freed automatically when data is deleted (or fup purgedata is done). You can release unused allocated extents with the ‘FUP Deallocate’ command.
What is Enscribe? A file system or a database management system?
Enscribe is NOT a file system. It is a hierarchical type of database management system. Tandem systems support two kinds of databases: NonStop SQL/MP which is a Relational Database Management system, and Enscribe.
“What does the MP suffixed in some of Tandem products names mean? E.g. NonStop SQL/MP, NonStop TM/MP, etc”
The MP stands for Massively Parallel and characterize products that have capabilities for
massively parallel processing.
What is the difference between NonStop TS/MP and Pathway/TS?
The NonStop TS/MP and Pathway/TS make up the OLTP environment for Compaq NonStop servers. The NonStop TS/MP
manages “processes that fall under the server category (of the requester-server model) like pathmon, pathcom,
server, server classes, linkmon, etc. The “Pathway/TS subsystem manages objects that fall in the requester
category like the TCP, TERM and PROGRAM objects.”
What is the difference between program and a process?
A program is a compiled executable code residing on the disk.
A process is a running entity created by executing the program; it is under the control of the operating system and it resides in the main memory of the processor. Processes compete with each other for the shared resource (RAM) while programs reside on the disk.
“process: A program that has been submitted to the operating system for execution, or a program that is currently running in the computer.”
program: A set of instructions that a computer is capable of executing.
What is the difference between a persistent process and a NonStop process pair?
Persistent processes are the ones that maintain no data of their own and need only to continue to execute. For such processes, it might be appropriate simply to ensure that the process gets started whenever it stops. A monitor process that periodically checks the process status can restart such a process.
NonStop processes are process pairs – one primary process and one backup process – that execute simultaneously on two different CPUs with continuous checkpointing between the primary and the backup process. The backup process may be an active backup process or a passive backup process.
What are the disk file types support on the Compaq NonStop servers?
“Since the NonStop servers treat everything as a file, including a device or a process, we will broadly divide the files on NonStop servers into disk files and non-disk files.”
Disk files can be SQL files or Enscribe files. You access Enscribe files using the Enscribe database
record manager. You access SQL files using the NonStop SQL/MP product.
Types of Enscribe Files:
The Enscribe database record manager provides access to and operations on Enscribe disk files.
The Enscribe software is an integral part of the operating system. It supports the following file types:
“Key-sequenced files, in which records are placed in ascending sequence based on a key field. The key field is a part of the record.”
“Relative files, in which records are stored at locations relative to the beginning of the file.”
“Entry-sequenced files, where records are appended to a file in the order they are written to the operating system.”
“Unstructured files, in which records are defined by the application. Records are written to and read from a file using relative byte addresses within the file.”
Other than NetBatch and NetBatch-Plus are there other scheduling programs available for the NonStop platform?
Yes. The MultiBatch product from Insider Technologies is a client-server (client on MS Windows) application for doing exactly that. You can see the MultiBatch white paper here and the comparison between the NetBatch and MultiBatch products here.
Can I READ my TMF audit trail files?
Did you know that your TMF audittrails in the system can be read? They can even be deleted? Surprised!
SNOOP is a utility provided by TMF to handle the TMF AUDITTRAILS. The following are some basic functions it does:
(a) Display info on the AUDIT TRAILS.
(b) Restore AUDITTRAILS from a tape to a disk.
(c) Delete an audittrail file.
(d) It can even act as a requester or server.
Well.. We hope you would have started to search for SNOOP in your system? Just look in the $SYSTEM.SYSnn subvolume.
From which version of the NSK operating system did the support for RISC processors begun?
From D40 onwards.
Why move from MIPS processors to Intel’s Itanium processor?
(1) Mass production of the Intel’s processors and hence the cost advantage
(2) 64-bit architecture
(3) Donno if SGI will continue to be around to support the processors for long, their business emphasis is
likely to change.
(4) Intel being the largest chip-maker will continue to function the way it is now.
(5) It provides a preferred customer status to Compaq now because most of HP’s server ranges run on Itanium
processors and now NonStop servers also run on Itanium. So Compaq/HP becomes the biggest corporate
customer for Intel’s Itanium chip.
Of which popular Unix shell is the OSS shell (OSH) an adaptation?
“Korn Shell, the default OSS shell prompt is $”
OSS is build on which standards?
What are the various components of the DSM (Distributed Systems Management)?
DSM/TC (Tape Catalog)
What is the USERID file I see in $SYSTEM.SYSTEM?
The file $SYSTEM.SYSTEM.USERID contains the username and passwords for all the users of the system. At the
time of logon, TACL process checks the entered values against the contents of this file.
What are the PAID and CAID in relation to a process?
The CAID (creator access ID) of a process is the user id of the user who initiated the creation of the process. Normally the PAID (process access ID) of a process is the same as its CAID except when the PROGID attribute is set for the program file of the process, in which case the PAID is set to the owner of the file and not the user id of the creator of the process.
What is a CSS subvolume?
CSS stands for Communication subsystem.
What is the file $system.sysnn.directry for?
This file stores the DP2 (disk process) directory. It contains the labels of all the files on that particular
volume. This file exists on all the volumes of all the systems and cannot be seen using TACL fileinfo command but can be seen with FUP INFO command. It is advisable not to mess around with this file.
What does the FUP RELOAD command do?
The FUP RELOAD command physically reorganizes a key-sequenced file or SQL objects (table or index only) while allowing shared read/write access to the file or object. A RELOAD operation improves the access time and use of space for a key-sequenced file or SQL object that has undergone a large number of insertions, deletions, and updates with length changes.
What is the A0CINFO file and what is its use?
A0CINFO file is a distribution subvolume (DSV) file that contains information about a product and each of its files, including product and file dependencies, how the files are used and where they are placed, and which type of processor the product runs on. Every product and interim product modification (IPM) to be manaed by the Distributed Systems Management/Software Configuration Manager (DSM/SCM) is distributed in a subvolume and that subvolume must contain the product’s A0CINFO file.
What is the difference between a batch process and an OLTP?
OLTP: A system whereby each transaction is processed immediately, without the delay of accumulating transactions into a batch.
Batch Processing: A system whereby business transactions are accumulated over a period of time and prepared for processing as a single unit or batch.
On-line or Real-time vs. Batch Processing:
Batch processing permits each day’s work to be accomplished with a cheaper machine. On-line processing reduces to a minimum the time people spend “just waiting around,” but only if the system hardware and software have so much extra capacity that they can meet the peak load, which means that there are many hours of the day hen the system is “loafing.”
With priority-based scheduling of the CPU (common in multi-user operating systems) the interactive users will get prompt attention whenever they need it, but the otherwise idle time for the CPU will instead be assigned to completion of the less urgent jobs submitted for batch processing.
How do I add or delete a user to the system?
Use the TACL ADDUSER and DELUSER commands if you do not have Safeguard running. Otherwise use the Safecom’s ADD USER and DELETE USER commands.
What is a disk cache?
Disk cache is the memory space that the disk process writes to before the data is actually written to the disk. It acts like a buffer.
What is TMF?
What is the process PIN?
Each process running on a system has a unique process ID (PID) that consists of two parts – first is the CPU number on which the process is running and the second is the Process Identification Number (PIN) which uniquely identifies a process within a processor.
How do I find out the PIN of a process?
A TACL STATUS <process-name | process number) will display the PID of the process or the PIDs of the primary and the backup process if the process is a part of a NonStop process pair.
What is the advantage of running a process as HIGHPIN?
How many processes can I run on a single processor?
What is native mode?
What happens during SYSGEN?
During SYSGEN the DSM/SCM tool takes in the SUT (Site Update Tape) and generates a system image in the form of $SYSTEM.SYSnn.OSIMAGE file. Optionally you may also create SIT (Site Image Tape) during sysgen which can be used to restore the system in case of a system crash.
What happens during COLDLOAD?
COLDLOAD is the process of booting up the system. During coldload you mention which OSIMAGE(i.e. OSIMAGE file from which $SYSTEM.SYSnn) you wish to use for booting up and the system boots up using that OSIMAGE. You can have OSIMAGEs from different SUTs (various releases of the operating system or with different components of the operating system) in various $SYSTEM.SYSnn’s and you can COLDLOAD (boot) using your choice of OSIMAGE and hence you can easily have any version of the NSK you wish.
Aren’t there any Ester Eggs on Tandem?
Not one that I know of!
There are however some undocumented tricks and shortcuts that are useful sometimes.
For example, NOFT’s “HELP UNDOCUMENTED” command gives information about some really useful but undocumented (and hence unsupported!) commands.
When do I use HIGHPIN attribute for my processes?
What is a big-endian system and a little-endian system of binary representation?
Big endian is the convention of numbering bits, bytes, or words in memory or in registers in a left-to-right order, so that the leftmost, most significant item has the smallest offset or smallest address. The entire object is addressed by the address of its biggest (rightmost) end. TNS/R systems are big endian.
Little endian is the convention of numbering bits, bytes, or words in memory or in registers in a right-to-left order, so that the rightmost, least-significant item has the smallest offset or smallest address. The entire object is addressed by the address of its little (rightmost) end.
Which of the above systems do Tandem systems follow?
TNS/R systems are big endian.
What is the difference between SQL/MP and SQL/MX?
Both products provide relational database management facility on NonStop servers.
SQL/MX is the latest version of SQL/MP product. The conversational interface to SQL/MX is called MXCI while that of SQL/MP is called SQLCI.
How many processors can I have on a single system?
Sixteen. If you want more think about Servernet clusters.
How many systems can I connect together in an EXPAND network?
255 systems numbered from 0 to 254.
How many systems can I connect together in a Servernet Cluster?
24 systems. Servernet cluster technology enables up to 24 servers to be connected in a group, or ServerNet cluster, that can pass information from one server to any other server in the cluster using the ServerNet protocol.
What is a super cluster?
When do I need to use the RUN command for executing a program and when do I need not use it? (implicit run and explicit run commands)
RUN command is an explicit run command required only for running files in the current subvolume when you have the file both in the current subvolume and also the #PMSEARCHLIST. When you know for sure that there is only one file with that name in (your current subvolume and #PMSEARCHLIST) you need not issue the RUN command.
In the following example I have a batchcom in junk subvolume and the system subvolume. To run the one in the present subvolume I need an explicit run command otherwise batchcom command will run the one in $system.system
$data003 junk 7> #pmsearchlist
#pmsearchlist expanded to:
$data003 junk 8> batchcom
BATCHCOM – T9190D30 – (30NOV2001^ACZ)
COPYRIGHT COMPAQ COMPUTER CORPORATION YEAR 1986-1999
$data003 junk 9> run batchcom
BATCHCOM – T9190D30 – (15APR2002^ADA)
COPYRIGHT COMPAQ COMPUTER CORPORATION YEAR 1986-2002
$data003 junk 10> fi batchcom
CODE EOF LAST MODIFIED OWNER RWEP PExt SExt
BATCHCOM 100 725996 20JUN2002 13:49 240,124 NUNU 74 16
$data003 junk 11> fi $system.system.batchcom
CODE EOF LAST MODIFIED OWNER RWEP PExt SExt
BATCHCOM 100 149504 09NOV2001 7:02 255,255 NUNU 74 16
$data003 junk 12>
Please note that this is applicable only if you do not have #defaults set in the #PMSEARCHLIST.
If #defaults and $system.system both appear in the #PMSEARCHLIST then run or no run command it will
always search in the subvolumes in the order in which they are mentioned in the #PMSEARCHLIST.
$data003 junk 12> #set #pmsearchlist $system.system #defaults
$data003 junk 13> batchcom
BATCHCOM – T9190D30 – (30NOV2001^ACZ)
COPYRIGHT COMPAQ COMPUTER CORPORATION YEAR 1986-1999
$data003 junk 14> run batchcom
BATCHCOM – T9190D30 – (15APR2002^ADA)
COPYRIGHT COMPAQ COMPUTER CORPORATION YEAR 1986-2002
In the above command the implicit run command defaults to the batchcom in $system.system while an explicit run command is required to run the one in the current subvolume….this is coz’ of the sequencing of the subvolumes in the #PMSEARCHLIST.
What is the difference between multiprogramming and multiprocessing?
multiprogramming: multiple programs resident in memory at any one time.
multiprocessing: multiple processors available at any one time
difference: one refers to programs in memory, the other to CPUs
Processes are necessary for multiprogramming as they provide a mechanism for OSes to encapsulate resources needed for a running program, monitor resource usage of those running programs, switch from a program blocked on an I/O to another which is able to use the CPU, create/suspend/destroy processes, etc.
What is a deadlock?
A deadlock is a situation in which two computer programs sharing the same resource are effectively preventing each other from accessing the resource, resulting in both programs ceasing to function.
The earliest computer operating systems ran only one program at a time. All of the resources of the system were available to this one program. Later, operating systems ran multiple programs at once, interleaving them.
Programs were required to specify in advance what resources they needed so that they could avoid conflicts with other programs running at the same time. Eventually some operating systems offered dynamic allocation of resources. Programs could request further allocations of resources after they had begun running. This led to the problem of the deadlock. Here is the simplest example:
Program 1 requests resource A and receives it.
Program 2 requests resource B and receives it.
Program 1 requests resource B and is queued up, pending the release of B.
Program 2 requests resource A and is queued up, pending the release of A.
Now neither program can proceed until the other program releases a resource. The operating system cannot know what action to take. At this point the only alternative is to abort (stop) one of the programs. Learning to deal with deadlocks had a major impact on the development of operating systems and the structure of databases. Data was structured and the order of requests was constrained in order to avoid creating deadlocks.
What is an OSS zombie process?
Same as the zombie process of the UNIX environment. A zombie process is a process that has finished executing and will be deleted at a later time.
Why do zombie processes exist? When a process finishes it calls the exit system routine. The kernel then releases all system resources previously used by the process, but saves the exit status and process ID. The process is now in the zombie process state. The process remains a zombie process until its parent process checks the exit status and tells the kernel to completely delete the process.
What are the advantages of having a message-based OS (e.g. NonStop Kernel) than a shared-memory OS (e.g. Windows 9X)
What are the advantages of having a RISC processor over a CISC processor?
CISC: Complex Instruction Set Computing. Most PC’s use CPU based on this architecture. For instance Intel and
AMD CPU’s are based on CISC architectures. Typically CISC chips have a large amount of different and complex
instructions. The philosophy behind it is that hardware is always faster than software, therefore one should make
a powerful instructionset, which provides programmers with assembly instructions to do a lot with short programs.
In common CISC chips are relatively slow (compared to RISC chips) per instruction, but use little (less than RISC)
RISC: Reduced Instruction Set Computing. RISC chips evolved around the mid-1980 as a reaction at CISC chips. The
philosophy behind it is that almost no one uses complex assembly language instructions as used by CISC, and people
mostly use compilers which never use complex instructions. Apple for instance uses RISC chips.
Therefore fewer, simpler and faster instructions would be better, than the large, complex and slower CISC
instructions. However, more instructions are needed to accomplish a task.
An other advantage of RISC is that – in theory – because of the more simple instructions, RISC chips require fewer
transistors, which makes them easier to design and cheaper to produce. Finally, it’s easier to write powerful
optimised compilers, since fewer instructions exist.
What really distinguishes RISC from CISC these days is more deeply rooted in the chip architectures.
*RISC microprocessors have more general-purpose registers. The best way to write fast code is to maximize the number
of operations performed “on-chip” and minimize the number of accesses to data stored in memory. More registers
make this goal more attainable. Register accesses are nearly instantaneous; memory accesses take time. A Pentium
has just eight general-purpose registers, but a PowerPC chip (used by Mac) has 32.
*RISC microprocessors use load/store architectures. CPU instructions that operate on data in memory-for example,
integer add instructions that add x to y, where y is a value stored in memory-are among the most expensive in terms
of clock cycles. In addition, the implementation of these instructions tends to require a disproportionate number
of transistors. RISC designs minimize the number of instructions that access memory in favor of load/store
architectures that require y to be loaded into a register before x is added, then “stored” back to memory.
*RISC microprocessors use uniform instruction lengths. On a Pentium, the length of one instruction can vary from as
little as 1 byte to as many as 7 (more if 32-bit code is executed in a 16-bit segment). RISC designers, on the
other hand, favor making all instructions the same length-usually 32 bits. This simplifies the instruction-fetching
and decoding logic and also means an entire instruction can be retrieved with one 32-bit memory access.
*RISC microprocessors emphasize floating-point performance. Traditionally, the stronghold of RISC machines has
been the scientific community, whose applications do more floating-point math than a typical word processor or
spreadsheet. Therefore, RISC microprocessors nearly always have high-performance floating-point units built in.
Only recently, with the Pentium, has an entire line of Intel chips included a math coprocessor (remember
the 486SX?), and integer operations are still considered the Pentium’s mainstay.
What is the difference between symmetric and asymmetric processing multiprocessing?
The difference between symmetric and asymmetric multiprocessing: all processors of symmetric
multiprocessing are peers; the relationship between processors of asymmetric multiprocessing is a master-
slave relationship. More specifically, each CPU in symmetric multiprocessing runs the same copy of the
OS, while in asymmetric multiprocessing, they split responsibilities typically, therefore each may have
specialized (different) software and roles.
The difference between symmetric and asymmetric multiprocessing maybe the result of either hardware or
software. Special hardware can be differentiate the multiple processors, or the software can be written to
allow only one master and multiple slaves. For instance Sun’s operating system SunOS Version 4 provides
asymmetric multiprocessing, whereas Version 5 (Solaris 2) is symmetric on the same hardware.
What are the advantages of DMA?
DMA is useful for transferring large quantities of data between memory and high-speed devices. It
eliminates the need for the CPU to be involved in the transfer, allowing the transfer to complete more
quickly and the CPU to perform other tasks concurrently.
What are the purposes system calls and system programs?
a. Purpose of system calls: System calls allow user-level processes to request services of the operating
b. Purpose of system programs: System programs can be thought of as bundles of useful system calls. They
provide basic functionality to users and so users do not need to write their own programs to solve common
What are virtual disks?
Virtual disk is a process name used instead of a physical volume name to provide location-independent naming.
Used by Compaq’s SMF (Storage Management Foundation) product.
51 How do I find out if a disk is a virtual disk?
SCF info disk $* command displays information about all the disks on the system. The disks with disk subtype 26
are all virtual disk.
Also FUP VOLS command shows all zeros in Capacity (total and free), free, count, biggest columns for the virtual
How do I add more services to the Telserv? (Besides TACL and Exit)
How come a NonStop server have more than one IP addresses?
What are the advantages of parallel library TCP/IP over the normal TCP/IP?
What are spin locks?
Spin locks are the fundamental and most common synchronization primitives used in in drivers.
Three types of spin locks are supported for DDI drivers:
(1) Simple (mutex) spin locks
Spin locks are the fundamental and most common locking primitive provided by DDI; they are also sometimes known
as simple locks, basic locks, or mutex locks. The context contending for the lock busy-waits until the lock is
acquired at the requested IPL. Spin locks are non-recursive and provide access to shared resources where the
context holding the lock cannot block. And in cases where the expected wait time does not warrant a context switch.
Spin locks can be used to guard critical code regions that are relatively long if there is low contention for
the resource protected by the lock. Spin locks must not be held when there is a possibility of a context switch.
(2) Read/write spin locks
Read/write spin locks are used when distinguishing the nature of access can enhance concurrency. Read locks permit
multiple concurrent lock holders, while write locks are exclusive, meaning that there can be at most one owner
of the lock in the write mode. If the lock can not be acquired immediately in the needed mode, the caller spins
until it can acquire the lock. Read/write spin locks are used when the lock is mostly acquired in the read-mode.
These locks have a strong writer preference, so a spinning writer will acquire (have preference for) the lock
before a spinning reader. This eliminates starvation for the contexts that want to acquire the lock in the
write-mode. As the expected access pattern is read-mostly, starvation of contexts that want to acquire the lock
in the read-mode is unlikely.
(3) Trylock spin locks
Trylocks are non-blocking lock operations; versions are supplied for both simple and read/write spin lock types
to use when blocking would be inappropriate.
How do I find out the IP address of the workstation on which I am working from the Tandem terminal emulator?
What do the characters in the file security string mean?
What is an alternate key file?
What is a table?
Definition obtained from http://searchdatabase.techtarget.com/sDefinition/0,,sid13_gci804740,00.html
In computer programming, a table is a data structure used to organize information, just as it is on paper.
There are many different types of computer-related tables, which work in a number of different ways. The following
are examples of the more common types.
I) In data processing, a table (also called an array) is a organized grouping of fields. Tables may store
relatively permanent data, or may be frequently updated. For example, a table contained in a disk volume is updated
when sectors are being written.
2) In a relational database, a table (sometimes called a file) organizes the information about a single topic into
rows and columns. For example, a database for a business would typically contain a table for customer information,
which would store customers’ account numbers, addresses, phone numbers, and so on as a series of columns. Each
single piece of data (such as the account number) is a field in the table. A column consists of all the entries
in a single field, such as the telephone numbers of all the customers. Fields, in turn, are organized as records,
which are complete sets of information (such as the set of information about a particular customer), each of which
comprises a row. The process of normalization determines how data will be most effectively organized into tables.
3)A decision table (often called a truth table), which can be computer-based or simply drawn up on paper, contains
a list of decisions and the criteria on which they are based. All possible situations for decisions should be
listed, and the action to take in each situation should be specified. A rudimentary example: For a traffic
intersection, the decision to proceed might be expressed as yes or no and the criteria might be the light is
red or the light is green.
A decision table can be inserted into a computer program to direct its processing according to decisions made
in different situations. Changes to the decision table are reflected in the program.
4)An HTML table is used to organize Web page elements spatially or to create a structure for data that is best
displayed in tabular form, such as lists or specifications.
What is an index?
What are protection views?
What is RDF/IMP and RDF/IMPX?
What are Nevada 4 and Nevada 4.5 systems?
S76000 and S86000 systems are also known as Nevada 4 and Nevada 4.5 systems respectively.
What is the command interpreter to RDF application?
RDFCOM command interpreter is provided to control overall RDF operations and is similar in functionality to TMFCOM,
the control utility that is used with TMF.
What is SNOOP? What application can be used to read the TMF audit trail files?
What do you mean by file index depth?
What is the performance implication of file index depth?
As file index depth increases the performance decreases.
What can be done to minimize file index depth?
What can I use to create a file on disk?
Many things. You can use TACL’s CREATE command to create an unstructured file on disk. You can use FUP’s CREATE
command to create almost any kind of Enscribe file with various configurable parameters like block size, max-extents,
primary and secondary extents, file code, etc.
You can also create files on disk programmatically using FILE_CREATE_ Guardian procedure call.
You can use SQL/MP or SQL/MX to create files on disk.
How do I enable/disable a volume for auditing by TMF?
What can I use to defragment a disk on Tandem systems?
You can use DCOM (Disk COMpression) product to defragment a disk and consolidate the free extents on a disk.
What is the functionality/features of DSAP?
The Disk Space Analysis Program (DSAP) analyzes how disk space is used on a specified volume. DSAP copies the disk
directory and free-space table to the current work file. By specifying options, you can manipulate this data to
produce several different reports about the use of the disk space for that volume. The free-space table is limited
only by your primary (main) and secondary (contiguous disk space) memory requirements.
How do I identify if a disk file on my Tandem system is corrupt or broken?
A C after the filename in a FUP INFO command output indicates that the file is corrupt.
That attribute of a corrupt file can be reset by using the FUP ALTER FILENAME, RESETCORRUPT command. It should,
however, be noted that this command does not affect the internal structure of the file and only resets the corrupt
flag for that particular file.
What do you mean by disk I/O congestion? What can be done to eliminate the same?
What is the difference between application throughput and system throughput?
What is throughput?
Why should I use Safeguard?
Because Safeguard provides additional facilities and functionalities that the standard Guardian security.
Safeguard features provide additional capabilities in the following areas:
(1) Authentication: More control is provided over authentication attempts and password management through global
configuration attributes and through individual user authentication records. Additionally, you can apply
special authentication features such as blind logon and password expiration warnings. You can also define user
aliases, which give users alternate names with which to log on to the system.
(2) Authorization: Protection of objects is extended to include volumes, subvolumes, devices, subdevices,
processes and subprocesses as well as disk files. Each object can be protected by an authorization record that
contains an access control list (ACL).
An ACL allows you to specify access to a much greater level of detail than you can with Guardian security strings.
(3) Auditing: Records can be generated for attempts to access an object as well as attempts to change the
authorization record associated with an object. You can also record attempts to log on as well as attempts to
change the user authentication records. Additionally, you can audit the actions performed by specific users.
(4) Administration: The task of security administration can be distributed among the security staff through
OBJECTTYPE authorization and membership in security groups. In addition, you can run other programs directly
from SAFECOM, the Safeguard command interpreter.
Certain security operations, such as the management of user aliases and file-sharing groups, are available only
through the Safeguard software. Further, all users who will work in the OSS environment must be added and managed
with SAFECOM USER commands.
File-sharing groups are supported only through the Safeguard software. A file-sharing group cannot be used for
managing user authentication records. Its purpose is to designate arbitrary groups of users who can share files,
especially in the OSS environment. A file-sharing group is defined with the ADD GROUP command. Members, who are
existing users, are added to and removed from a file-sharing group with ADD GROUP and ALTER GROUP commands.
With SAFECOM GROUP commands, you can make a user a member of multiple groups for file-sharing purposes, even
though that user can have only one administrative group. A single user or user alias can be made a member of up to
32 groups. The names of all groups to which a user belongs are retained in the user’s group list. With these
commands, you can also add a user to other administrative groups solely for file-sharing purposes. This allows
larger groups of users to share files because you can expand an administrative group beyond the 256 members
to which it is limited for user administration. File-sharing groups can be particularly useful in the OSS
File-sharing group names and numbers cannot appear on a Safeguard access control list, but they can be used in
the OSS environment to specify group IDs for file permission codes. In addition, the Guardian G and C file
security codes recognize all groups in a user’s group list. The Safeguard software also recognizes a user’s
group list when it evaluates an access control list entry that specifies all members of a particular group.
For example, the entry 141,* on an access control list grants access to all members of group 141, including
those users who are file-sharing members of the group.
What encryption does TACL use to encrypt the data kept in the $SYSTEM.SYSTEM.USERID file?
When you enable the Safeguard PASSWORD-ENCRYPT configuration attribute, passwords are encrypted using DES(Data
Encryption Standard) as a one-way encryption algorithm. The system can verify passwords but cannot decrypt them.
Passwords are stored in encrypted form in the USERID file. When a user supplies a password to log on, it is
encrypted and compared to the value in the user ID file. If they match, the logon attempt is successful.
Can I find the username and password back from $SYSTEM.SYSTEM.USERID file?
No. Because the passwords are encrypted using DES as a one-way encryption algorithm. So only encryption possible
no decryption possible.
How do I define an ALIAS for my username in Safeguard? What are the advantages?
What is used to install OSS on a system? I understand that DSM/SCM as yet does not install OSS installation?
Is there a chance that in future DSM/SCM can be used for OSS installation?
What is a block?
The block size is the number of bytes that are transferred between the disk and the disk process. The block size
can be 512, 1024, 2048, or 4096 bytes. Records cannot span blocks; therefore the block size must be at least
large enough to contain one record and the overhead associated with the block. In other words, the maximum record
size is smaller than the block size. A block usually contains multiple records.
What do you mean by extractors and updators when talking about RDF?
What does GARTH stand for? What can it be used for?
Why do we need AutoTMF product? What does it do?
What does ATALLA products do? What is their USP?
Why should we prefer hardware based encryption?
Does the iTP web server support SSL protocol?
What other security techniques does iTP web server support for tranmitting data over internet?
What are Active Transaction Pages?
What activities can be audited by Safeguard?
How do I determine if the file is corrupt or not? How can a file get corrupt?
> FUP INFO * WHERE NOT CORRUPT
The file can get corrupt if the DUP or LOAD operation is completed abnormally.
FUP INFO command displayes a C after the filename to indicate that the file is corrupt.
What information can you get from FUP for OSS files?
FUP can only handle OSS files with the FUP INFO command. OSS files cannot be
specified in any other FUP commands.
What are SQL views? What types of SQL views are supported by NonStop SQL product?
What can I do recover the corrupt file?
> FUP ALTER FILENAME, RESETCORRUPT
What is the default value for EXTENTS for Enscribe files?
By default every Enscribe disk file and partition has 16 extents; you can specify that a
particular nonpartitioned file or the partitions of a key-sequenced file have more than
What file codes I cannot use for my files? Is there a default file code if I do not specify one?
File codes 100 through 999 are reserved; using them causes unpredictable results.
“If you do not specify the file code, it defaults to 0.”
Ok I know what are Format 1 and Format 2 files. What NSK version do I need if I want support for Format 2 files?
The D46 release supports larger partitions then were previously supported. This change
allows files and file partitions to increase from the format 1 size of 2 gigabytes to the
format 2 size of 1 terabyte (although the upper limit of the actual file size depends upon
the size of the largest single disk).
What is an Application Architecture?
I keep hearing about the CPSA marketing Ids for NonStop software products. What is CPSA?
CPSA (Customer Process System Architecture) is Compaq’s intenal tool used to customer support people
worldwide that can be used for various purposes including ordering products, quoting,
order entry and management, fulfilment of orders, and also has an accounting interface.
What is the role of EPO (Emergency Power Off)?
EPO is a set of equipments installed at a site that allows the disconnection of power to ALL the equipment in that
region in cases of emergency. An EPO connector on the service side of an enclosure allows an external signal to
disable the batteries to the NonStop. EPO should be used only in emergency situations.
Is there a certification track to test and demonstrate my skill level for Compaq NonStop platform?
Describe the importance of ESD protection?
Can I install a NonStop Himalaya system in a very dusty environment?
Search for particle in TIM or NTL to look for air particle limitations.
What is the SCF command to find out all the virtual disks on a system?
> SCF STATUS DISK $*, SUB VIRTUAL
What is the subtype for virtual disks?
36. For all disks the device type is 3 but subdevice type vary depending on the type of the disk.
Why do I sometimes need a raised floor for installing NonStop systems?
NonStop Himalaya computer systems that are installed in a dedicated computer room must always be placed on a
raised-floor structure. In addition, NonStop Himalaya computer systems consisting of two or more rows of
enclosures that are installed outside of a computer room must be placed on a raised-floor structure.
The primary reason for having a raised floor is that power cables and data cables can then be installed on the
subfloor, underneath the raised floor. Such placement eliminates the safety and damage risks posed when cables
are placed on an open floor, where they might be stepped on frequently.
Which products of K-series did TSM replace in S-series systems?
RMI, SYSHEALTH, and TMDS
109 Since which operating system version do we have support for RISC processors on NonStop systems?
What is $ZHOME?
What is VHS?
What is an Inspect Monitor process?
What processors do the current NonStop systems use?
What function do I use to compare two strings in TACL?
Use #MATCH in-built function. #MATCH compares two strings irrespective of their cases.
Demonstration of the usage of #MATCH
$data00 william 14> #match william “william”
#match william “william” expanded to:
$data00 william 15> #match william william
#match william william expanded to:
$data00 william 16>
$data00 william 16> #push temp
$data00 william 17> #set temp william
$data00 william 18> #match temp william
#match temp william expanded to:
$data00 william 19> #match [temp] william
#match william william expanded to:
$data00 william 20> #match william [temp]
#match william william expanded to:
$data00 william 21> #match William [temp]
#match William william expanded to:
$data00 william 22> #match WILLIAM [temp]
#match WILLIAM william expanded to:
Why do some products have T numbers (numbers beginning with a T) while some products have
S numbers (numbers beginning with S)
What is the advantage of using QIO?
What do I do to view the status of all the UP CPUs on a system?
How do I restrict one particular user from using FTP on my NonStop server?
You can restrict users to FTP to S-series server by creating FTPUSERS file in $SYSTEM.ZTCPIP subvolume.
Just create this file and mention the name of users that u want to restrict.
How do I access OSS files on a NonStop server while connected using FTP?
Can Backup/Restore product set be used to backup OSS files also?
No. You can backup and restore OSS files using pax utility on OSS. Also you can backup OSS files using pax
and then copy the pax file to the Guardian name space, and then use Backup/Restore.
While I am connected to a NonStop server by FTP how do I find out which file system is currently active?
What is dynamic process configuration?
Dynamic process configuration: Using Subsystem Control Facility (SCF) to configure a generic process to always
start in a designated primary processor (that is, to be fault-tolerant).
Since what version of the NSK is HIOGHPIN supported?
Where does NonStop servers use SEEPROM?
What are the advantages of using SEEPROM?
Why do you need dual fans for cooling purposes?
What is the purpose of disk cache?
What happens to my running processes if the power supply to the NonStop servers is cut off abruptly?
The batteries in the server will keep it running for some time (at least 2 hours) and then the memory of the
processors will be frozen till the time the power supply is restored.
When the power supply is indeed restored the execution starts exactly from where it left. And also the batteries
start charging themselves.
How does worm hole routing improve Servernet router performance?
Wormhole routing is for lowest-latency switching.
ServerNet technology uses a technique known as wormhole routing to reduce network latency. With this technique,
a packet does not need to be completely received before being sent to its next destination (as with store and
forward). Wormhole routing works by allowing the router to decode the header of the packet as it is received,
then locate the port on which the packet will exit by using the header’s destination address and the internal
routing table. This one-time operation allows the packet to be directed through the crossbar switch while the
remainder of the packet is still being received by the router. As a result, the header is routed and retransmitted
well before the end of the packet has been received. Thus, the latency incurred by the router is much less than
that of store-and-forward technology.
How does a requester-server architecture differ from a client-server architecture?
The client-server architecture evolved from the original requester-server architecture. In a requester-server
architecture both the requester(s) and server(s) reside on the same platform and they communicate using a common
protocol. An example of a requester-server application is a Pathway application running SCOBOL requesters and
communicating with Pathway servers…both the requester and server are on the same (Compaq NonStop) platform).
In a client-server architecture, the client(s) and requester(s) reside on different platforms and communicate using
a common protocol. An example is DSM/SCM where the client is residing on a PC (Windows on Intel/compatible platform)
and the server is residing on a Compaq NonStop server (server is a NonStop TS/MP server class).
Another example is TSM (Tandem Service Management) application.
What is the difference between the CONVERSATIONAL (CONV) mode and BLOCK mode of the T6530 terminal?
What is an asynchronous terminal?
How do you calculate availability?
Availability is normally calculated as the percentage ratio of the uptime to the total time
(i.e. uptime + repair time).
Assuming a system with 1 hour of downtime in a duration of 100 hours, the availability
will be 99/100 * 100 = 99 %
How do I measure downtime in a client-server application?
Normally outage is measured in user-outage minutes.
Normally calculated as the number of minutes the application is unavailable multiplied by the number of affected
What is the difference between FTP and TFTP protocols?
FTP (File Transfer Protocol) uses TCP(OSI layer 4-Transport layer-protocol) over IP (OSI layer 3-Network layer-
protocol while TFTP (Trivial FTP) uses UDP(Another OSI layer 4 protocol) over IP protocol for transmission.
It is to be recalled here that TCP is a connection oriented protocol while UDP is an unreliable connection-less
What are the standard TCP/IP port numbers used by the operating system and not available to the users?
Port numbers which are fixed for some services are not available for programming/configuration by the user are
known as well-known port numbers. The well-known port numbers for Compaq NonStop Himalaya systems are listed in the
file $SYSTEM.ZTCPIP.SERVICES. The following is a dump of a typical SERVICES file.
# @(#)services 1.16 90/01/03 SMI
# Network services, Internet style
# This file is never consulted when the NIS are running
tcpmux 1/tcp # rfc-1078
discard 9/tcp sink null
discard 9/udp sink null
systat 11/tcp users
chargen 19/tcp ttytst source
chargen 19/udp ttytst source
smtp 25/tcp mail
time 37/tcp timserver
time 37/udp timserver
name 42/udp nameserver
whois 43/tcp nicname # usually to sri-nic
hostnames 101/tcp hostname # usually to sri-nic
piccolo 2787/udp piccolo # Cornerstone Software
piccolo 2787/tcp piccolo # Cornerstone Software
# Host specific functions
link 87/tcp ttylink
x400 103/tcp # ISO Mail
pop2 109/tcp # Post Office
nntp 119/tcp usenet # Network News Transfer
ntp 123/tcp # Network Time Protocol
NeWS 144/tcp news # Window System
# UNIX specific services
# these are NOT officially assigned
shell 514/tcp cmd # no passwords used
printer 515/tcp spooler # line printer spooler
courier 530/tcp rpc # experimental
uucp 540/tcp uucpd # uucp daemon
biff 512/udp comsat
who 513/udp whod
route 520/udp router routed
new-rwho 550/udp new-who # experimental
rmonitor 560/udp rmonitord # experimental
monitor 561/udp # experimental
pcserver 600/tcp # ECD Integrated PC board srvr
ASM 1961/tcp #ASM User Interface
NSOADMIN 2222/TCP #ODBC
KELVIN 2220/TCP #ODBC
phiwsock 1089/tcp # DSM/SCM Planner Interface
What is the definition of a transaction?
How do I find out who developed a product e.g. who developed TMF? What if I wish to meet the developer face-to-face?
You can get yourself registered in the Yahoo groups Tandem_Alumni and post your query. Many of the really old Tandem folks are registered and visit frequently at this group. You can also contact your GCSC and may be he knows the person you want to meet! Besides the Compaq NonStop magazine “24 X 7” also runs articles occasionally about the original developers of some popular products.
Are there any ester-eggs on Compaq NonStop (Tandem for us) servers?
What is DSM (Distributed Systems Management)?
DSM is a set of software applications, tools, and services that facilitates management of subsystems, networks,
and business applications.
What is an SRL (Shared run-time library)?
Shared run-time libraries (SRLs), sometimes called shared resource libraries, are object files used by more than
one process at a time. SRLs are usually used by system programs to reduce memory usage or replace problematic
code with corrected code, but you can create an SRL for a TNS/R native application so that more than one process
can run the same program module.
What is an Action Event in the even messages reported to the operator by EMS?
An event that requires operator intervention to resolve.
What is the function of the process $NCP?
The network control process ($NCP) is responsible for initiating and terminating server-to-server connections
and maintaining network-related system tables, including routing information. $NCP must be running at every node
in the Expand network before Expand lines can be started.
What is the X.25 protocol? Where is it used?
What is HDLC(High-level data link control) protocol? Where is it used?
What is ATM (Asynchronous transfer protocol)? Where is it used?
What do the terms bps and Baud refer to when talking about modems?
What is a Servernet cluster?
ServerNet Clusters use Expand to provide a high-speed interconnect between servers over a limited geographic
range. Three network topologies are supported: the star, split-star, and tri-star topologies. The star topology
supports up to eight nodes. The split-star topology supports up to 16 nodes. The tri-star topology supports up
to 24 nodes.
What is a Ferris Wheel topology?
The Ferris wheel topology is recommended when a ServerNet cluster needs to coexist with a FOX ring.
What is a virtual circuit? What kinds of Virtual circuits are there?
A virtual circuit is a circuit or path between points in a network that appears to be a discrete, physical
path but is actually a managed pool of circuit resources from which specific circuits are allocated as needed
to meet traffic requirements.
A permanent virtual circuit (PVC) is a virtual circuit that is permanently available to the user just as
though it were a dedicated or leased line continuously reserved for that user. A switched virtual circuit (SVC)
is a virtual circuit in which a connection session is set up for a user only for the duration of a connection.
PVCs are an important feature of frame relay networks and SVCs are proposed for later inclusion.
Permanent Virtual Circuit:
A permanent virtual circuit (PVC) is a software-defined logical connection in a network such as a frame relay
network. A feature of frame relay that makes it a highly flexible network technology is that users (companies
or clients of network providers) can define logical connections and required bandwidth between end points and
let the frame relay network technology worry about how the physical network is used to achieve the defined
connections and manage the traffic. In frame relay, the end points and a stated bandwidth called a Committed
Information Rate (CIR) constitute a PVC, which is defined to the frame relay network devices. The bandwidth may
not exceed the possible physical bandwidth. Typically, multiple PVCs share the same physical paths at the same
time. To manage the variation in bandwidth requirements expressed in the CIRs, the frame relay devices use
a technique called statistical multiplexing.
Switched Virtual Circuit:
SVC also stands for supervisor call.
In a network, a switched virtual circuit (SVC) is a temporary virtual circuit that is established and maintained
only for the duration of a data transfer session. A permanent virtual circuit (PVC) is a continuously dedicated
virtual circuit. A virtual circuit is one that appears to be a discrete, physical circuit available only to the
user but that is actually a shared pool of circuit resources used to support multiple users as they require the
connections. Switched virtual circuits are part of an X.25 network. Conceptually, they can also be implemented as
part of a frame relay network.
What is the difference between TCP and UDP protocols?
How does the TCP/IP protocol suite map to the layers of the OSI reference model?
What is the difference between RJ-45 and RJ-11 connectors? Are there any other RJ connectors?
How do I find out what is the contents of a file from its file code?
How do I find out the index level of a SQL object?
SQLCI FILEINFO command can be used to determine index-levels and extent and data block
use of a table or index.
What impact does index level have on the performance?
Performance increases if the index levels are less.
Using OSS functions can I read/write to Guardian files?
Not all. Only odd-unstructured Guardian files can be accesses from OSS functions.
No structured files can be accesses.
Can OSS processes be run as a process pair?
What is the difference between a MOM process and an ancestor process?
If the created process is unnamed then the creator process is called the MOM process of the
created process. If the created process is named then the creator process is calle the ancestor
process of the created process.
What is the difference between symmetric and asymmetric encryption?
What are the common/popular symmetric and asymmetric encryption techniques?
What are the $DMnn process on my system?
How do I create a backup of my TACL process?
#backupcpu n command
Can I interchange the primary and backup cpus for my TACL process? How?
What is the difference between waited and no-wait I/O?
How do you create saveabend files?
TACL>#set saveabend on
What is the advantage of isolating data from application?
Known as data encapsulation, minimizes the code that must change to support a change in data layout.
What are the ZYQ files on my system?
They are OSS files. The files are present in the subvolume name beginning with ZYQ and the filenames begin with
Z0000 usually. FUP INFO command displays the filecode of these files as OSS. And their security strings (RWEP) is
UNIX security string (drwerwerwe). The output of TACL FILEINFO command will be different with the file code being
100 and RWEP string ####.
What do you mean by ACID properties of transactions?
ACID- The basic properties of a database transaction: Atomicity, Consistency, Isolation, and Durability. All
transactions should comply with these properties.
Atomicity – The entire sequence of actions must be either completed or aborted. The transaction cannot be
Consistency – The transaction takes the resources(database tables) from one consistent state to another.
Isolation – A transaction’s effect is not visible to other transactions until the transaction is committed.
Durability – Changes made by the committed transaction are permanent and must survive system failure.
How do I find out which version of the NSK OS is running on my system?
How does the backup process come to know that the primary process has failed?
A “process deletion” system message from the OS to bakcup process tells that the primary process has failed
and a “processor down” message from the OS to the backup process tells that the CPU on which the primary
process was running is down or is otherwise inaccessible.
The receipt of process deletion message is automatic while for receiving “processor down” message the
backup process must call the MONITORCPUS procedure.
What is the primary key for a database table?
A primary key uniquely identifies a row ofdata found iwthin a table.
What is MXCI?
It is the conversational interface to the SQL/MX product. Just like SQLCI for SQL/MP.
What components of Java does NonStop Server for Java (NSJ) – Tandem’s implementation of Java- NOT support?
AWT and AppletViewer due to nonvisual nature of Tandem servers.
What are Java Beans?
NSJ 1.5 includes Tandem’s implementation of Java Beans. Java Beans are reusable s/w components that can run
in both a design environment (inside a builder tool) and a run-time environment.
How do I tell TAL compiler where it should allocate memory for the data in each variable?
(1) Variable declaration with no indirection symbol = memory allocated in the user data area
(2) Variable declaration with . Indirection symbol = memory allocated in the user data segment using indirect
(3) Variable declaration with .EXT indirection symbol = memory allocated in the extended data segment.
How do I find out which is the current EMS log file?
What are Codd’s rules for database design?
What is normalization in databases?
What is Nagle’s algorithm used in all terminal emulators including Tandem’s T6530 terminal emulator?
Named for its creator, John Nagle, the Nagle algorithm is used to automatically concatenate a number of small
buffer messages; this process (called nagling) increases the efficiency of a network application system by
decreasing the number of packets that must be sent. Nagle’s algorithm, defined in 1984 as Ford Aerospace and
Communications Corporation Congestion Control in IP/TCP Internetworks (IETF RFC 896) was originally designed to
relieve congestion for a private TCP/IP network operated by Ford, but has since been broadly deployed.
Nagle’s document specified a means of dealing with what he called the small packet problem, created when an
application generates data one byte at a time, causing the network to be overloaded with packets (a situation
often referred to as send-side silly window syndrome). A single character – one byte of data – originating from
a keyboard could result in the transmission of a 41 byte packet consisting of one byte of useful information
and 40 bytes of header data. This situation translates into 4000% overhead, which was considered to be acceptable
for a lightly loaded network such as the Advanced Research Projects Agency Network (ARPANET – the precursor of the
Internet operating at that time), but not so for a heavily loaded network such as Ford’s, where it could necessitate
retransmissions, cause lost packets, and hamper propagation speed through excessive congestion in switching nodes
and gateways. Throughput could be hampered to the extent that connections were aborted. Nagle’s algorithm – usually
implemented through the insertion of two lines of code into a TCP program – instructs the sender to buffer (store)
data if any unacknowledged data is outstanding. Any data sent subsequently is held until the outstanding data is
acknowledged (ACKed) or until there is a full packet’s worth of data to send.
In short, “Nagle’s algorithm or heuristics are used in better TCP/IP implementations for traffic reduction”.
Is there any cable length limitation for RS-232 connections?
What classes of IP addresses are there?
From an IP address how do I find out what class the network falls in?
What is the factory default IP addresses of MSP0 and MSP1 of a NonStop Himalaya server?
What are routing tables?
In a Servernet network which component is responsible for the generation of routing tables?
Why does PING command not work on my NonStop server?
What is the function of DNS?
What are the major SNA protocol families supported by NonStop servers?
What are the IEEE standards related to the Ethernet and Token Ring communication standards?
What is Action View 400?
How do you find out the firmware level in a component of NonStop server?
How do I update the firmware version of a component of NonStop server?
How do I identify the firmware files on my NonStop server?
They are code 510 files.
What are the contents of the CSS subvolume in my volume $SYSTEM?
CSS stands for communication subsystem
What utility do I use to examine network data packets on a NonStop server?
Do we have any industry recognized certification for demonstrating skills on the NonStop platform?
What are the benefits of NonStop Himalaya certification?
What is the function of the file $SYSTEM.ZTCPIP.HOSTS?
Helps TCP/IP in host name to IP address resolution.
Are there any products that can help me give my S-series server a web server functionality?
How do I disable certain ports on my NonStop server? E.g. to disable the FTP connections to my system
In what kinds of network does X.25 normally operate?
“It is generally used in packet switched networks of common carriers, such as telephone companies.”
What does the Compaq NonStop Himalaya Cluster switch do?
A hardware product that controls the routing of packets across the external X and Y fabrics of a ServerNet cluster.
What does the ECHO command do?
$system william 8> echo terra
Establishing Connection to 192.168.97.25…Connected
a quick brown fox jumps right over a bad dog
a quick brown fox jumps right over a bad dog
$system william 12> echo /in taclcstm/ terra
What is iTP ATP?
206 What is TCP/IP multicast and broadcast?
How do I configure a TCP/IP subnet address on a NonStop server?
What are the various types of nodes on a SNA network?
What is the function of socket interface on NonStop systems?
What are the primary subsystem processes on a NonStop system?
What is a Concentrator Manager process?
When do I use to Compaq WebSafe2 ISP products?
Compaq Websafe2 ISP Product Description:
“Uncompromising security. Unrivaled speed. For safe, assured Internet commerce.”
Features at a glance
•Provides unsurpassed application-level cryptographic services
•Supports public and symmetric key cryptography
•Increases security and reliability with automated key management
•Delivers industry-leading physical and logical security
•Provides secure hardware encryption for securing sensitive data
•Offers easy administration and control
•Export version available
What do I use to configure the QIO environment on my system?
What are independent products?
What is the function of the PAM subsystem?
What configuration files are needed for proper configuration of NonStop TCP/IP?
In which types of networks are the ATM protocol used?
How do I see the MAC address of an E4SA?
info pif $zzlan.e401.*
What are MAC addresses? What is their importance?
How do I start a TCP/IP process?
What is the function of Bulk I/O
How do I monitor attempts to logon to my server using a particular user id?
What is the medium through which two enclosures interconnect?
What is enclosure interleaving?
What is the function of CIIN file at system startup?
How do I write to the line 25 of the T6530 terminal screen?
How do I use tandem spooler to print to the network laster printer?
I use setprompt both to see the volume-name and subvolume name on the TACL prompt. Are there any other options
for the SETPROMPT command?
How do I connect to other people who use Tandem? Is there any user’s group for Tandem?
What is ATLAS? I hear of it a lot in job requirement advertisements!
I compile a SCOBOL program and it does not produce an executable code (file with filecode 100 or 700).
There are 2 or 3 files that are generated. How do I run a SCOBOL program then?
Maximum how many jobs can I have in a spooler at a time?
What are the advantages of using spooler plus over usign spooler?
What are the advantages of using NetBatch-Plus over NetBatch?
What is the difference between an IPM and an NPV?
The terminology has now changed to Successive PVU and Initial PVU respectively
What is the largest implementation of Compaq NonStop servers….with maximum number of processors?
What is the biggest advantage/feature of the Pathway/iTS product?
Allows you to convert your SCOBOL requester programs to web clients i.e. html pages
What is a system area network (SAN)?
What is ServerNet System Area Network (SAN)? What are its advantages?
How do I open 2 files simultaneously in TEDIT and switch between them as I do with CTRL+TAB of MS Word?
What are the 14 files that are generated when I create a DDL dictionary?
What kinds of files are the 14 files that are generated when I create a DDL dictionary?
All the files except DICTDDF are key-sequenced; DICTDDF is an unstructured file.
Why do I see TACL inbuilt functions when I press ESC on my PC when I am working in Outside View?
That is because, by default, the ESC key is mapped to F16 – the TACL help key.
This can be verified by using #HELPKEY command on TACL prompt.
This can be changed by using #SET #HELPKEY F12 (or any other function key) on the TACL prompt.
Still the ESC key will continue to be mapped to the F16 key but since F16 is no longer the TACL help key the TACL builtins won’t be displayed on pressing the ESC key.
How can I map the function keys on my keyboard to perform specific functions on a TACL prompt?
You said that data travels as data packets over a ServerNet fabric. What is the size of a ServerNet packet?
What is a ServerNet adapter?
A ServerNet adapter provides the interface between a ServerNet fabric and an I/O bus such as a SCSI bus. A ServerNet adapter contains a ServerNet bus interface (SBI) and one or more ServerNet addressable controllers (SACs).
What do you mean by worm-hole routing?
Original source: http://nonstop.compaq.com/view.asp?IO=SNETHSPD
ServerNet technology uses a technique known as wormhole routing to reduce network latency. With this technique, a packet does not need to be completely received before being sent to its next destination (as with store and forward). Wormhole routing works by allowing the router to decode the header of the packet as it is received, then locate the port on which the packet will exit by using the header’s destination address and the internal routing table. This one-time operation allows the packet to be directed through the crossbar switch while the remainder of the packet is still being received by the router. As a result, the header is routed and retransmitted well before the end of the packet has been received. Thus, the latency incurred by the router is much less than that of store-and-forward technology.
What are the first systems to implement the ServerNet architecture?
S7000 and S70000
What is the difference between an S7000 and an S70000?
What is an SEB?
SEBs connect the ServerNet X and Y fabrics of a processor enclosure to the ServerNet fabrics of another system enclosure, allowing all the processors and peripheral devices in your system to communicate over the ServerNet fabrics. In a processor enclosure, slots 51 and 52 are generally reserved for SEBs. If these slots are not used for SEBs, they must contain either ServerNet adapters or filler panels to maintain proper air flow and electromagnetic interference (EMI) shielding.
What is a ServerNet fabric?
Can I compile a TAL or C program on my PC and them copy it to Tandem server and make it run there?
Using a tool called DCE.
“While working on my terminal can I send messages to other users on their terminal similar to the UNIX “wall” command?”
What is the advantage of ETK (enterprise tool kit)?
Why should I use the CA Unicentre product for NonStop systems?
What editors does OSS support?
What is the function of the viewsys product?
What is the use of NFS (Network file system?
I have partitioned the hard disk on my PC. Can I also partition the disk on a Tandem server?
What is the function of the pathmaker product?
What is a TLE (Time-list-entry)? I see a TLE field in the output of PEEK command?
Why do I need a raised floor for installing a NonStop Himalaya system/
What are processor halt codes? How do I find out the halt code if the processor is halted?
Ok. I can send mails from Tandem. Can I also read my Exchange server mails from Tandem?
Why should I migrate my TAL program to pTAL?
How do I prohibit other users from reading/writing to my subvolume?
Can I also use Safeguard to secure my processes?
Why use optical disks? Advantages?
What is SNMP?
What is a Nomadic disk?
What does SWID do?
Where can I find the SYSGEN reports? Names of the files?
What are the contents of the CONFTEXT file? What if I change them?
I heard I will be able to compile my programs on my PC using Microsoft Visual Studio .NET and copy the program to Tandem and it will run. Is it true?
Yes. The Enterprise Toolkit product of Compaq is fully integrated with MS Visual Studio .NET and comes as an extension package to the same. It supports programming in native C, native C++, pTAL and native COBOL languages and the compilation can be done using Visual Studio .NET and the compiled code is then FTP’ed to NonStop server and executed there. It has enhanced SQL, CORBA and TUXEDO support and is scheduled to ship in Q3, 2002.
It also works with ODBC/MX to access SQL/MX database online and also privides design time database access.
What is CRE (common run-time environment)?
What software product do I need to use DSM/SCM?
What is the function/advantage of burse detection and suppression feature of EMS?
How do I view the EMS messages that are currently being generated on my PC’s terminal?
Is mixed language programming possible on Tandem systems?
Yes. You can do mixed language programming in TAL, C, COBOL85, FORTRAN and Pascal. That is, you can call C procs from Tal program and vice versa.
What are labelled tapes? How do they differ from unlabelled tapes?
Is there a third party product available with the same functionality as the Compaq RDF product?
Yes. Golden Gate’s Extractor/Replicator product does the same thing.
What is the use of the Tandem Reload Analyzer product?
What is a context-free server?
What is a context-sensitive server?
What is a sensitive command? Can you give me some examples of sensitive commands?
How do I list all the devices available on my system?
How do I get the kernel subsystem information using the SCF product?
How do I get the TCP/IP subsystem information using the SCF product?
How do I get the storage subsystem information using the SCF product?
How do I get the LAN subsystem information using the SCF product?
How do I get the WAN subsystem information using the SCF product?
How do I communicate with the Nomadic disk subsystem?
using SWCI command interface.
If I want to learn about NonStop Himalaya systems from scratch from TIM where do I start?
Start from reading the Himalaya S-series Server Description Manual in TIM
How many internal disks can be accommodated in an enclosure of an S-series system?
What is the size of a word on a NonStop system?
A RISC word is four bytes wide and most subsystems in the NonStop Himalaya S-series
“servers use this word width as a normal convention. However, the TNS instruction set,”
“originally used in the earliest Tandem (now Compaq) systems, assumes a two-byte”
What is big-endian system as opposed to a little-endian system? In which category does Tandem systems fall?
Big endian: The convention of numbering bits, bytes, or words in memory or in registers in a left-to-right order, so that the leftmost, most significant item has the smallest offset or smallest address. The entire object is addressed by the address of its biggest (rightmost) end. TNS/R systems are big endian.
Little endian: The convention of numbering bits, bytes, or words in memory or in registers in a right-to-left order, so that the rightmost, least-significant item has the smallest offset or smallest address. The entire object is addressed by the address of its “little” (rightmost) end.
What is the size of a memory page on a NonStop system?
“A unitary segment is 128 kilobytes, or eight pages.”
“A page is 16,384 bytes of information. When stored in physical memory, a page”
“occupies one 16,384-byte physical frame of storage space.”
“memory page. A unit of virtual storage. In TNS systems, a memory page contains 2048”
“bytes. In TNS/R systems, the page size is determined by the memory manager and can”
“vary, depending on the processor type.”
What is meant by absolute addressing?
Absolute addressing: A context-independent mode of addressing that addresses elements in one of the absolute segments. For TNS/R processors, the absolute segments are segments 8192 through 16381, which are in the Kseg2 portion of virtual memory.
Absolute addressing can be used only in privileged mode.
What is relative addressing?
Relative addressing. A mode of virtual addressing in a processor that each process can use to address its own user space portion of virtual memory, relative to page 0 of segment 0.
What is the ALLPROCESSORS paragraph in the CONFTEXT file?
ALLPROCESSORS paragraph: A required paragraph in the CONFTEXT configuration file that contains attributes defining the Compaq NonStop Kernel operating system image for all system processors. The ALLPROCESORS paragraph follows the optional DEFINES paragraph.
What is an AWAN access server?
Asynchronous Wide Area Network (AWAN) access server: A local area network (LAN)-based communications device that provides
(1) asynchronous connections to terminals, printers, and terminal emulators for Compaq NonStop Himalaya S-series and K-series servers;
(2) remote access for disk operating system (DOS), Microsoft Windows, and Macintosh computers;
(3) VT-to-6530 protocol conversion; and
(4) dial-out connections for LAN-attached DOS, Windows, and Macintosh computers.
What is a firmware?
Firmware: Code in memory that is necessary for the power-up initialization and communication with a host or device. The software for components of the ServerNet architecture (for example, an adapter) is called firmware. Some firmware for ServerNet components is downloaded when the system or component is loaded.
What is an interrupt? What is an interrupt packet?
Interrupt packet: A ServerNet packet containing a ServerNet address that has been set up as an interrupt address.
The access validation and translation (AVT) logic recognizes this address as special, and sends the packet data to the tail of an interrupt queue.
What is a memory manager?
Memory manager: A Compaq NonStop Kernel operating system process that implements the paging scheme for virtual memory. This process services requests generated by different interrupt handlers as well as by other system processes.
Memory manager process exists on a per processor basis. It runs as a single process and not as a NonStop process pair.
What is a page fault?
Page fault: A reference to a logical page that is not currently in physical memory. Such a reference causes an interrupt, and the interrupt handler begins a sequence of operations that loads the page into memory.
What is a persistence count?
Persistence count: The number of times the $ZPM persistence manager process will restart a generic process that has been terminated abnormally. A generic process with an AUTORESTART value of 10 (the maximum) is said to have a persistence count of 10.
What is a persistence manager process?
Persistence manager process: The $ZPM process that is started and managed by the $ZCNF configuration utility process and that starts generic processes in G-series releases and manages their persistence.
Tell me something about the founder of Tandem and its initial days.
OR the local copy
I understand there are two kinds of enclosures on S-series systems – system enclosures and peripheral enclosures.
Can you give me one example of a peripheral enclosure?
The modular storage system that contains 519x tape CRUs and other peripheral devices is an example of peripheral enclosure.
What is a Group? How do I find out the Group number?
A group consists of all objects accessible to a pair of SP in a system enclosure. In an S-series system a group comprises of all the components in a system enclosure.
The group number of an enclosure is indicated on the faceplate of the door of the system enclosure.
What is a module? How do I find out the module number?
A module is a subset of a group and consists of components sharing a common hardware interconnection. A module contains one or more slots. In an S-series server there is exactly one module in a group (and hence the module number always 01. Module number is shown on a label located inside of the enclosure near the fans.
What is a slot? How do I find out the slot number?
A slot is a physical, labelled space in a module. Slot number is indicated next to its slot.
I have a multi-CPU system with many enclosures with separate power supplies. How then am I able to power on my system using a single switch?
PMF CRUs in all the system enclosures of a system are connected using 4-wire power-on signal cables with RJ-11 connectors, allowing the system to be powered on from a single power on push-button.
How do I power on a Himalaya S-series system?
The system can be powered on by pressing the power-on push-button on either of the PMF CRUs in group 01. The power-on
signal is carried from the PMF CRUs to all the other PMF CRUs (using the power on signal cable – RJ11 connectors)
in the system thus powering on the entire system.
How do I power off a Himalaya S-series system?
Power off facility is not provided by the power-on signal cable (RJ11). Power off capability is provided by
a s/w command. In emergency, you can unplug the AC power cords from the wall, cut off the power at the circuit breaker or use the EPO switch (if installed). Care must be exercised if using the EPO switch since it even disconnects the battery from the system and hence the system experiences abrupt power failure – to be used only in emergency purposes.
Why am I advised against using EPO switch for powering off the system?
Because the EPO signal, if connected, will even disconnect the battery supply to the NonStop system and hence the system will experience an abrupt power failure.
What are group ID switches? Where are they located?
Group ID switches are located in slots 24 and 27 – two for fault tolerance – on the appearance side of a system enclosure. These should match with each other and with the group ID mentioned on the enclosure door faceplate.
How many physical disks does a mirrored volume contain?
Two disks installed in slots that are on different SCSI buses.
I know that the slots numbered 09 and 19 terminate SCSI bus 1 and slots numbered 10 and 20 terminate SCSI bus 2. What is the function of SCSI bus terminators? Why are they needed?
About half the trick to SCSI is getting the termination right, and 90 percent of what’s left is being sure
you don’t have cable problems. –Jerry Pournelle
On a SCSI bus, you must have two terminators, one at each end of the bus. That is, if all your devices are
internal, you should validate the termination on the adapter and plug a terminator on the last physical device
(last physical device, and not last logical address).
Conversely, if you only have external devices, the termination should be validated on the adapter and a terminator
should be plugged in on the last device. There are also devices whose termination is enabled/disabled by switches.
Details about SCSI topology:
When setting up a SCSI bus, it is imperative that each device be in a straight line. This means that each device
(including the host adapter) is connected to either one or two other devices, and never more than that.
The two devices at the end of the bus must be terminated, either internally or externally. The bus should never
be connected in a loop, star or other formation.
For a two-device bus, the topology typically looks like this:
Terminator — Device A — Device B — Terminator
For a four-device bus, it is as follows:
Terminator — Device A — Device B — Device C — Device D — Terminator
A good article on SCSI bus termination is available at this URL:
A local copy can be found at:
What kind of batteries are used in the Himalaya S-series servers?
Lead-acid batteries with 24 sealed, cylinderical cells rated at 2.5 Ampere-hours each. Batteries installed in a
system keep recharging themselves and can power the memory for a minimum of 2 hours in case of power failure.
How does the system console connect to the S-series Himalaya system?
System consoles connect to the system through an ethernet connector labelled “ETHERNET” on the PMF CRU for the group 01 enclosure. This port is unused in PMF CRUs of other group enclosures.
How does an MSP differ from an SP?
Service processors (SP) are always installed on MFIOB (Multi-function I/O boards) in PMF CRUs. If that PMF CRU is installed in group 01 then this SP is called MSP (Master Service Processor). This MSP connects to the system console through the “ETHERNET” port on the PMF CRU.
SPs located in other groups are called ESPs (Expansion Service Processors).
What kinds / types are there for the NonStop Himalaya S-series servers?
NSR-W = S7000
NSR-D = S7400
NSR-G = S70000
NSR-T = S72000
NSR-V = S74000
You can remember the sequence using the phrase: “When Didya Go To Venice?” And take the initials of all words and you have the genesis of S-series systems…NSR-W to NSR-V.
You can use peek/cpu_no./ command to find the NSR number of the CPU and then, from the above table, find out the exact CPU model number.
What is ‘nslookup’ ?
nslookup is the name of a program that lets an Internet server administrator or any computer user enter a host name (for example, “geocities.com”) and find out the corresponding IP address. It will also do reverse name lookup and find the host name for an IP address you specify.
An example of running nslookup on MS DOS prompt:
Default Server: adm.bhmi.tcpn.com
Addresses: 220.127.116.11, 18.104.22.168
How do I start and stop Safeguard?
You can start Safeguard by running the file OSMP in $SYSTEM.SYSnn.
TACL> RUN $SYSTEM.SYS01.OSMP/NAME $ZSMP, CPU 0/1
And then go to Safecom and issue “START SAFEGUARD” command.
To stop Safeguard just go to Safecom prompt and issue “STOP SAFEGUARD” command.
What is Extended Memory Segment?
Is there any interprocess message advantage (e.g. faster) within a CPU?
On D-series and subsequent systems, yes. But on C-series systems there is no interprocess message performance
advantage within a CPU. (I do not know why. Do mail me if you know.)
What is $CMON?
Original source: http://www.overlord.com/v5.1/Bal-CMON.html
$CMON is an optional process, written in any programming language, to which Tandem’s command interpreters (TACL and SQLCI) pass certain requests (for approval or modification) before they pass these requests to the operating system. There is no standard $CMON; each Tandem system may or may not have a $CMON written specifically for that system or copied from another system.
Each TACL process opens the $CMON process the first time that it receives a command that causes a request to be
sent to $CMON. If, for any reason, the TACL process is unable to open the $CMON process, then the
command-interpreter process goes ahead and processes the command as if $CMON had accepted the request.
$CMON programs can monitor and control requests to the command interpreter and the Safeguard subsystem, such as logon requests and explicit or implicit RUN commands. For example, you could write a $CMON program that does not allow certain individuals to log on during nonbusiness hours or from certain terminal locations. $CMON programs can also require the use of user names (instead of user IDs) and blind passwords during logon attempts.
$CMON programs monitor requests only to the command interpreter or the Safeguard subsystem. Requests made through other programs, such as TEDIT, are not monitored. Protect the object code for $CMON programs so that only the owner has WRITE authority.
What do you mean by CPU queue length?
The length of time processes spent ready and waiting for CPU cycles to be spent on their behalf, divided by the total elapsed time in the relevant sample. A queue length of zero means no processes have had to wait for the CPU to be available; a queue length of 2.5 means that, on average during the sample period, between two and three processes were waiting. (There could easily have been more than three at times and less than two at other times.) Queues are bad. A queue length near 1 or 2 is considered “OK” on a busy system, but the closer to zero the better.
What do you mean by dispatches? What is dispatch rate?
Dispatch is the number of times a process was selected from the ready list and executed by the CPU.
Dispatch rate is the number of dispatches per second. A dispatch occurs when a CPU changes context from one executing process to another.
What are high PCBs and low PCBs?
Each process in each CPU occupies a PCB. Low PCBs are those numbered less than 255; High PCBs are those numbered greater than 255.
What is a DCT (Destination Control Table)?
Destination Control Table (DCT) is a collection of operating system data structures that serves as a directory of named processes and logical devices. It is a collection of PCBs.
In addition to a PCB associated with every process, the operating system maintains several other tables of information to keep track of processes. A collection of such tables, known as the destination control table (DCT), contains information about all named processes on the system. This table is a system-wide table and therefore remains visible even if a processor should fail.
What is a PCB (Process Control Block?
Process Control Block (PCB) is an operating system data structure that contains information about the resources and environment of a process. Each process has a PCB. The process creation attempt by your program may fail if there are no more process control blocks available.
A process control block (PCB) is used by the operating system to control process execution. The PCB contains pointers to the process code and data areas, retains process context when the process is suspended, and contains pointers to files opened by the process.
For example, a saveabend file is created for the process if the saveabend bit is set for the process in the
process control block (PCB).
What does a process consist of?
A process consists of the following:
· Code areas in virtual memory that contain the instruction codes to be executed. These code areas are shared by all processes in the same processor that execute the same program file. The instructions in the code areas in virtual memory are derived from the code part of the program file on disk.
· Data areas in virtual memory that contain the program variables and temporary storage that is private to the process. Even if other processes use the same code areas, each process has its own private data areas. The disk part of the data area is obtained from the Kernel-Managed Swap Facility (KMSF), or, in some cases, from a designated swap file.
· A process control block (PCB) that is used by the operating system to control process execution. The PCB contains pointers to the process code and data areas, retains process context when the process is suspended, and contains pointers to files opened by the process.
What are Cyclone machines?
Cyclone servers represent an important milestone in the development of the present HP NonStop servers.
The first and second generation Tandems were called Tandem Non-Stops (TNS) and TNS-IIs. TNS and TNS-II models gave way to TXP, VLX, Cyclone and, most recently, Himalaya models.
[Detailed information about the design and features of the Cyclone machines available in the following Tandem Technical report – http://historical.ncstrl.org/tr/fulltext/tr/tamucs/TR95-007.txt%5D
What is the relation between RDF and TM/MP (TMF) products?
If your database is audited, TMF (a.k.a. TM/MP) guarantees that every file update has an audit trail record. In conjunction with TMF online dumps, the audit trail can be used to recover lost files up to the state of the last committed update.
RDF uses the TMF audit trails to reliably and efficiently replicate the database to a second copy on a remote Tandem system. RDF guarantees that all audited updates are quickly transmitted to the remote system and applied to the remote copy of the database.
For serious Tandem-to-Tandem replication-and for proven value in disaster recovery situations-there is no more efficient, reliable, or complete solution than TMF/RDF.
How do I change/remove the AUDIT attribute of a disk file?
$data00 temp 18> fup alter wlog0001, audit <<to give the AUDIT attribute
$data00 temp 18> fup alter wlog0001, no audit <<to remove the AUDIT attribute
How do I change/remove the PROGID attribute of a disk file?
$data00 temp 35> fup secure job,,progid << to give PROGID attribute
You can dup the file to the same/another location; this removes its PROGID attribute.
What is power-fail delay time?
During an AC power failure, the batteries can continue to power the components in a system enclosure for a certain period of time. After that period of time, the batteries continue to power the processor memory for a minimum of two hours. The amound of time that batteries can continue to operate all components in an enclosure is called the power-fail delay time.
It is specified as POWERFAIL_DELAY_TIME attribute of the SCF kernel subsystem.
How do Himalaya S-series servers connect to the public LANs?
Using E4SA(Ethernet 4-port servernet adapter) Ethernet ports that are not being used for connecting to SWAN concentrators.
What TIM manual should I start with? I am a new Tandem system administrator.
Introduction to NonStop Operations Management.
What is a memory-exact-point?
Memory-exact point is a location in an accelerated program at which the values in memory (but not necessarily in the register stack) are the same as they would be if the program were running on a TNS processor. Most source statement boundaries are memory-exact points. Complex statements might contain several such points:
at each function call, privileged instruction, and embedded assignment.
What is a register-exact-point?
Register-exact point is a location in an accelerated program at which the values in both memory and the register stack are the same as they would be if the program were running on a TNS processor. A register-exact point is also a memory-exact point.
What is QUICC?
Quad-integrated communications controller (QUICC). The Motorola MC68360 chip. For Compaq NonStop Himalaya S-series servers, the QUICC is used as the service processor (SP) and is the main part of the ServerNet wide area network (SWAN) concentrator communications line interface processor (CLIP).
What are the major advantages and disadvantages of satellite communications?
The satellite-connect line-handler process implements the satellite-efficient version of the HDLC protocol, HDLC Extended Mode. HDLC Extended Mode allows a maximum window size of 61 frames (the maximum window size is the number of outstanding frames that can be sent before an acknowledgment is required) and implements the selective-reject feature. Selective reject causes only frames that arrive in error to be retransmitted.
The major benefits of satellite connections are
· Price-to-performance ratio. Satellite channels can add a large amount of transmission capacity, significantly reducing the cost of long-distance communications.
· Fault-tolerance. You can use the multiline path feature to enhance the reliability of satellite connections.
Using this feature, you can configure up to eight parallel lines between nodes.
The major disadvantage of satellite connections is potentially long propagation delays (approximately 240 mlliseconds) when sending data to the satellite and then to the destination node. The reliability of satellite connections can also be adversely affected by weather and other atmospheric conditions.
What were the driving forces in the development of the ATM protocol?
Asynchronous Transfer Mode (ATM) technology is based on the efforts of the International Telecommunication Union Telecommunication Standardization Sector (ITU-T) Study Group XVIII to develop Broadband Integrated Services Digital Network (BISDN) for the high-speed transfer of voice, video, and data through public networks.
How do I determine the product version (VPROC) of DSM/SCM?
> Volume $dsmscm_vol.ZDSMSCM
> Vproc taexe
> Vproc cbexe
What is the difference between user managed swap space and kernel managed swap space?
The Kernel-Managed Swap Facility (KMSF) is the preferred method for managing virtual memory using swap files for all D4x releases after D40.00 and for G-series releases. You need to configure each processor with at least one kernel-managed swap file that provides the swap facility needed by its processes. You can load your system using the default swap files, but you should also configure additional swap files. You do not want to risk halting your entire system when you run out of swap space.
Is there any recommended sequence of starting up the applications/subsystems and shutting down the system?
You can implement the system startup sequence with a collection of startup files, each with a specific purpose.
Compaq recommends that you invoke the startup files in this order:
1. Startup file for the system, to be invoked after the CIIN file is invoked
2. Startup files for the system software
3. Startup files for the subsystems
4. Startup files for the communications lines
5. Startup files for the applications
You can implement the system shutdown sequence with a collection of shutdown files, each with a specific purpose.
Compaq recommends that you invoke the shutdown files in this order:
1. Shutdown files for the applications
2. Shutdown files for the communications lines
3. Shutdown files for the subsystems
4. Shutdown files for the system software
5. Shutdown file for the system
Is there a way in which I can restrict/limit the powers of the super id?
Strictly speaking, there is no way you can restrict or limit the powers of the super id. The control you place on using the super ID depends on the importance given to security in your organization. The following methods will help you in managing/controlling the super id.
Some suggestions mentioned here might not necessarily be appropriate for all installations. You can limit knowledge of the super ID password to the security administrator and to the one or two people who handle emergencies. Change the password frequently, and keep the only written copy under lock and key. Where greater security is needed, take additional measures such as the ones that follow.
Dual Custody of the Super ID Password The security administrator can give half of the password to one person and the other half to another. Then when an emergency arises, two users are required to log on as the super ID. Accountability for any subsequent actions should be shared by these two users.
How much performance advantage can I expect when I add a processor to the NonStop Himalaya S-series system?
Himalaya servers, by virtue of their scalability feature, protects the investments of its customers by allowing them to keep adding processors and I/O devices and storage devices and gaining proportional gain in performance and capacity. The scalability feature of the Himalaya servers ensure that with each processor added to the system you gain at least 98.2% of the performance of the first processor in the system.
How does Compaq “fail-fast” principle contribute to the reliability of the NonStop Himalaya S-series servers?
System processes and critical hardware modules are designed to be fail-fast. In other words, they must perform to specified standards or they halt and go offline before any problem has the chance to propagate to other modules. Hardware and software are made fail-fast through extensive error checking. Some hardware components also perform periodic self-tests. The operating system performs rigorous internal consistency checks to verify its inputs, outputs, and data structures. In the extremely rare instance where an error occurs within a system process or the operating system detects a corrupted data structure, the operating system halts the processor and lets the backup processes in other processors take over. No two processors have identical states so the error condition is not repeated in the backup. This way, no malfunctioning system process is allowed to continue after the error is detected. Other vendors that do not support process pairs cannot react to failures in this way. In those cases, the operating system tries to continue rather than use the fail-fast technique. Such systems are vulnerable to data integrity problems and error propagation.
fail-fast: The technique of stopping a module immediately when a nonrecoverable error condition occurs, to ensure that the error condition is not propagated to other modules.
If I accidentally delete the super id is there any way I can recover it?
If the super ID is deleted from the system, there are ways to recover it. If no CIIN file was specified when the system was generated, you can perform a system load from the Operations and Service Processor (OSP).
The OSP operator becomes the super ID and can then add the super ID to the USERID file. If a CIIN file was specified when the system was generated, you must perform a system load from a tape. The USERID file on the tape contains an entry for the super ID.
How do I disallow anonymous logons in FTP / OSS on HP NonStop servers?
To logon to the OSS environment, the anonymous user enters the user name, “anonymous” or “ftp”. The FTP server checks this logon against the contents of the FTPUSERS file residing on $SYSTEM.ZTCPIP. If the logon is found in this file, the anonymous user logon is rejected.
What is the difference between TMF version 1, 2 and 3? What improvements / changes were done in the later versions?