Load NW Server 100% occasionally

Occasionally , one per week
utilization server is up to 100%
In shedulling information in monitor
no processes with big load, not more 5%
but work in last string about 100%

what is this perhaps ?
whatcan big load server ?
tolyannAsked:
Who is Participating?
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

pdeleonCommented:
What version of Netware and what is the hardware configuration of your server?
0
danilopCommented:
If the utilization is 100% only occasionally and not for a long time
it isn't a problem.
Or your question is about the fact that the sum of all % of all process
don't give 100%?

BYE DAN
0
tolyannAuthor Commented:
NW 4.11 sp5
server HP LHpro 2xPPro 200 Mhz 128 RAM
RAID 5 HPRaidController 8 GB

I need solution it's trobble
because users are very dissatisfied.
0
Cloud Class® Course: MCSA MCSE Windows Server 2012

This course teaches how to install and configure Windows Server 2012 R2.  It is the first step on your path to becoming a Microsoft Certified Solutions Expert (MCSE).

danilopCommented:
Try to read this TID about high utilization on NW4.
Then you can also read the TID 2941108 that contains a very
helpful document about high utilization (HIGHUTL1.EXE).

Troubleshooting High Utilization NW 4 Summary (Last modified: 23JUL1999)  TID2916430
This document (2916430) is provided subject to the disclaimer at the end of this document.
Issue
FYI:  This document is duplicated inside of highutl1.exe on the minimum patch list page of Novell's Support Connection web site.  It is suggested that highutl1.exe be used instead of this document because it contains some files that are helpful in troubleshooting and resolving high utilization issues.  In addition, this TID is just a summary whereas highutl1.exe contains a more thorough explanation of the items listed here.


TROUBLESHOOTING HIGH UTILIZATION FOR NOVELL NETWARE 4.1X


This document is also used for performance enhancement of NetWare server.


After each of these steps answer the question, "Is utilization still high", or "Has performance improved"?


1.  IS IT NORMAL??
2.  Is the high utilization number cosmetic?  Loading or unloading any NLM will cause processor information in monitor to be recalculated.
3.  Upgrade to NetWare 4.1x from a NetWare 4.0x server.
4.  Use the current OS patches.
5.  Use the current disk and lan drivers.
6.  Type "speed" at the system console to verify the processor speed.
7.  Make the server 'vanilla'; in other words remove all nlms except disk and lan drivers.
8.  Check in monitor.nlm | processor utilization | <F3> for the busiest thread on the server.  Unload it if it is a known nlm.
9.  Server resources such as min/max packet receive buffers or min/max directory cache buffers can cause utilization problems if they are set improperly. See TID 2905856 for more details.
10.  Use workthread.nlm (included in HIGHUTL1.EXE).
11.  Make sure the set parameter "upgrade low priority threads" is at its default ... OFF (except with SMP).
12.  NetWare 4.11 volumes with more than 2.5 to 3 million directory entries can experience utilization issues in specific circumstances.
13.  Upgrade the disk driver.
14.  Have 10%-20% free disk space - especially on the SYS volume - plus at least 1,000 free blocks on every volume.
15.  If suballocation is being used on server volumes, 64 KB is the recommended block size.
16.  If NetWare is decompressing large files, utilization will be high.
17.  Make sure that no directories are flagged IC (immediate compress).
18.  To check compression activity, type "set compress screen = on" at the system console prompt.
19.  "set deleted files compression option = 2" will cause high utilization.  Don't do this.
20.  File compression should occur after hours.  Make sure that the default values for the compression parameters have not changed.  In most high utilization troubleshooting issue, we suggest that compression is disabled until the issue is solved.
21.  Watch 'dirty cache buffers' and 'current disk requests' to see if they are high.
22.  Check the LRU sitting time in 'cache utilization' inside monitor.nlm - if consistently less than 15 minutes the server is short of RAM.
23.  Check cache buffers in monitor.nlm.
24.  Check all of the cache statistics in 'cache utilization inside monitor.nlm (use <F1> for help on individual statistics and their definitions).
25.  "Set reserved buffers below 16 meg = 200" should be in the startup.ncf file if the server has a SCSI or IDE device.
26.  Update the ds.nlm to the latest available.
27.  Directory services by default will have one master and two read/write replicas.  More replicas than this can lead to excessive traffic on the wire.
28.  Turn DS syncing off by typing this at the system console prompt:
               set dstrace = !D60
This command disables outbound syncing on the server where it is typed.
               set dstrace = !E
This command enable syncing to take place.  The number behind the 'D' is minutes.  So the 60 means that outbound syncing is diabled for 60 minutes.
29.  Watch the dstrace screen for syncing errors:
               set dstrace = on
               set dstrace = +s
               set dstrace = *h
30.  Bindery emulation can cause utilization issues.  If users connect to the server with netx, or if printers are bindery based they use one thread to process requests on the processor.
31.  Use no more than 40 printers per server, and make sure that all JetDirect and Netport calls to the server are being made through NDS, not the bindery.
32.  Check the print queues for corruption.  Delete corrupted queues.
33.  Type "set bindery context =" at the system console prompt.  If utilization drops, there is a bindery connection or program that is causing the utilization.
34.  Remote.nlm and Rspx.nlm will not cause high utilization.
35.  Check for hung connections and clear them.
36.  Update clients.  Run the latest client from Novell (or Microsoft) because known issues are fixed.
37.  Change the following NCP set parameters to ON from their default of OFF:
               display NCP bad component warnings
               reject NCP packets with bad components
               display NCP bad length warning
               reject NCP packets with bad lengths
38.  Some utilities use a "flat bindery scan", like syscon.  This can cause high utilization.  If flat bindery scans are the issue, stop using the program that causes them.  Or take a trace to find out where the scan is being originated from.
39.  Update the lan drivers to the latest available from the vendor.  Check for discepancies between the ODI version of the driver and the ODI version of the ethertsm/tokentsm and then msm modules.
40.  Turn packet burst off at the server.
41.  Use Category V cabling.
42.  Unplug the server from the lan.  See if the utilization drops.
43.  Check connections to see if an RJ-45 plug has been disconnected.
44.  Type "track on" at the system console prompt to see if traffic is excessive.
45.  Check the "no ECB" count in servman.nlm.  If it is incrementing then server memory, lan driver, or the lan card could be the issue, or a shortage of Service Processes or Packet Receive Buffers.
46.  Type "load spxconfg i=2400 q=1" at the system console prompt to increase available IPX sockets from 1200 to 2400.
47.  If appletalk is used on the server, update all of the nlms.  There is a field test appletalk.nlm that seems to be resolving many issues.
48.  If rconsole is run from a DOS box on Win 3.x/95/98/NT and made a background process, utilization can rise dramatically on the server.
49.  Use the 'Scheduling information' in monitor to reschedule a process's priority.
50.  Virus check the DOS partition.
51.  Mounting of volumes on NetWare servers will cause high utilization (CD or normal volume).  And sometimes a corrupt CD index will cause high utilization.  Type 'cd purge' at the console prompt to remove all of the CD indexes.


Extra Tips:


1.  Baseline the server and know the range of normal operation.
2.  What changed from the time when the server worked correctly, and when it didn't work correctly.
3.  Isolate the problem and make it reproducible.
4.  Use server -ns, server -na, server -ndb (or a combination of the 3) to help troubleshoot.
5.  Check server interrupts for conflicts, sharing, etc.
6.  Use the Novell 386 debugger to help identify processes that are causing high utilization or put offending threads to sleep.
7.  'Set maximum physical packet size' to the default for the topology used (i.e., 1514 for ethernet, 4202 for token ring).
8.  Check the support connection web site for other tids, or use the book "Novell's Guide to Resolving Critical Server Issues" for further troubleshooting.
9.  Mark with a P (Purge Immediate) flag directories which have a large number of temporary files created, such as print queue and email message directories. This will slow down the accumulation of dleted files using up space (see point 14 above).


Clarification: Setting the P flag on a directory does mean that files in that directory will be purged immediately, even though the files themselves do not get explicitly marked with the P flag. To quote from the NetWare 4.1 Concepts manual, under Directory Attributes: "Purge (P): Tells NetWare to purge the directory and any file in the directory when deleted, even if the purge attribute isn't set on an individual file. The directory and files can't be salvaged with the FILER utility".


Primus

Document Title:      Troubleshooting High Utilization NW 4 Summary
Document ID:      2916430
Document Revision:      17
Creation Date:      06NOV1996
Modified Date:      23JUL1999
Novell Product Class:      Connectivity Products
      Groupware
      Management Products
      NetWare
      Novell BorderManager Services
      Novell Directory Services

Novell Product and Version:      Dial-In/Dial-Out Connectivity
      NetWare MultiProtocol Router
      GroupWise
      ManageWise
      NetWare 3.11
      NetWare 3.12
      NetWare 4.0
      NetWare 4.01
      NetWare 4.02
      NetWare 4.1
      NetWare 4.11
      NetWare 4.2
      NetWare 5
      Novell Clients
      Novell Replication Services
      Novell Small Business Suite 5
      Z.E.N.works (April 98)
      Z.E.N.works 1.1
      intraNetWare 4.11
      intraNetWare for Small Business 4.11
      BorderManager 2.1 (also called 1.0)
      NDS for NT 1.0
      NDS for NT 2.0
0
IceyCommented:
If your users are doing alot of writes to the volumes included in your RAID array then the RAID 5 Array could be the culprit. While great for fault tolerance, it's not great on performance especially when you do alot of writes.
0
vernonlsCommented:
If this is only happening once a week then some weekly event is what is triggering it.  Could it be a tape backup job scheduled to run once a week?  Also, VirusScan programs are often set to run once a week.  

Is the time set correctly on the server?  If it is set incorrectly, the server may be trying to do file compression (normally set to run at midnight) during your work day.  That can cause very high utilization.

Like many things, adding memory to this server will also help.  going to 256MBytes will boost its performance, though mostly on reads from the server.

Also, there is a great Performance tuning document written for Backup Exec which Veritas emailed to me.  I would be happy to forward it to you if you send your email address to me.  It is some 40 odd pages, so too lengthy to post here.  Email me at
VernonLS@community.net

Hope this helps.

Vern
0

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
tolyannAuthor Commented:
Thanks all!!
0
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Novell Netware

From novice to tech pro — start learning today.