nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

iwextattr executes very slow

hemant_mcafee

Hi

We have two CMS boxes with TS6.5 SP2. At one box we facing performance issue but the same code works fine on other box. After doing some analysis, found that iwextattr command is executing very slow...ratio is

At one box it parse ~3 files per sec and ~20 files per sec on other...

I am just wondering what can be the cause for this, is it something related to iwperl???

Just doubt...

Hemant Goel

Find more posts tagged with

Comments

Adam Stoller

Hi

We have two CMS boxes with TS6.5 SP2. At one box we facing performance issue but the same code works fine on other box. After doing some analysis, found that iwextattr command is executing very slow...ratio is

At one box it parse ~3 files per sec and ~20 files per sec on other...

I am just wondering what can be the cause for this, is it something related to iwperl???

Just doubt...

Hemant Goel

My guess is there are other differences on the systems - CPU / memory / disk space / processes running, etc.

Any of these could affect the performance you are seeing.

iwextattr is a binary program and not a perl script - and thus iwperl, in and of itself, is unlikely to be the culprit.

mdang

Also, the invocation time of any CLT is faster on Solaris than on Windows, and faster on Linux than on Solaris, in case you are comparing between different platforms.

--Michael

hemant_mcafee

Thanks Guys

System configuration is almost same, OS is Solaris 9

Where performance is fine
-------------------------
System = SunOS
Node = 21cnedc23
Release = 5.9
KernelID = Generic_118558-14
Machine = sun4u
NumCPU = 4

Where performance is issue
--------------------------
System = SunOS
Node = cbu072a013
Release = 5.9
KernelID = Generic_118558-36
Machine = sun4u
NumCPU = 4

Hemant Goel

Nikunj

Are you sure it is iwextattr clt only...did you try with other clts..
Check your server...is der any background processes running.??....
and r u running this clt from an ipl or command line...??..If from ipl try running it from command line...

Migrateduser

Hi Nikunj

I tried both the options...both are slow Smiley Sad

You are right here,,, there are other cmds as well which are slow comparatively...

Hemant Goel

nipper

Check your networking setup.

try a few pings

ping localhost
ping server
ping server.domain.com

on both

Also, peruse the iw.cfg for any other hostnames lists (might be IP address).

Migrateduser

I checked the number of processes and also mount points...and found the difference,,,but wondering if it may be the cause...

Where performance is an issue:

Number of process running: 159
Usin NFS mount

Where performance is fine:

Number of process running: 89
Usin UFS mount

Hemant Goel

nipper

You can run ps -ef and see if any processes are using a lot of CPU, also worth a look to see if there are any perf issues (RAM, disk, etc) vmstat or some of the newer GUI tools will help.

Also, for grins try hard coding the path, /usr/iw-home/bin/iwextattr and see if that has any issues.

Nikunj

Ya nipper has given good suggestion.
I remember once our server(it is on windows 2000 server)slowed down. When i checked the processes from task manager it showed iwperl which took whole cpu memory.
This was because my OD got hanged; some one had initiated some bad OD deployment.
I stopped all Interwoven services especially OD services..and killed those iwplerl processes..And restarted all services using iwreset -a..
Hope this helps you to figure out the culprit... Smiley Happy

Migrateduser

Hi All

I checked the processes and cpu utilisation but all are fine it seems.

Hemant

nipper

is your networking correct ? Did you try the various ping commands ?

Migrateduser

Yes

Tried to contact with the Unix team as well and they also confirmed that CPU usage low on both its CPUs. Network traffic and NFS reads/writes look absolutely fine and the filer isn't showing any bottlenecks in writing to disk or offloading from cache.

As of now we are taking all the chances...but not sure whats the exact issue

Hemant

nipper

Take a look at iw.cfg and see if there are any strange looking entries (hostname, ip etc) in there

chicka

Yes

Tried to contact with the Unix team as well and they also confirmed that CPU usage low on both its CPUs. Network traffic and NFS reads/writes look absolutely fine and the filer isn't showing any bottlenecks in writing to disk or offloading from cache.

As of now we are taking all the chances...but not sure whats the exact issue

Hemant

Please have a check on the system cronjobs as well as try executing iwstat CLT recursively whcich will give you better anaysis report.

Migrateduser

Please have a check on the system cronjobs as well as try executing iwstat CLT recursively whcich will give you better anaysis report.

Sure Varun

Migrateduser

Hi All

Just wanna update that after changing the mount point properties, we can see 20-30% improvment...

Hemant Goel

PD007

Hi Hemant,
can you please tell me how can we change the mount point properties of server.
because i am also facing same issue.
Please suggest any way.