on 02-05-2010 8:27 AM
Hi Experts,
I am very new to NW admin side and currently my client is facing an issue while accessing portal via browser.
Problem: Some times suddently when users try to access production portal it takes a lot of time to open up and in most cases it doesn' t get opened up at all. To check what' s wrong with the portal i access sapmmc and inside their i find under current status -> ... -> services -> http provider -> total requests in red with a description as avg request - response time last reported value above threshold limit. It will be the same case under open alerts too.
My server configuration is:
SAP-JEE = 7.0SP14
OS = Windows 2003 (X86) 5.2
Server Version = 1.4.2_17-bo6
Kernel Version = 7.00 patch level 109886.44
Server = 32 bit
RAM = 16.00GB
Processor = Intel(R) Xeon(R) CPU E5450 @3.00GHZ (8CPUs)
Heap Size: 1024 MB
Approximate number of users = 400 - 500.
Currently to solve the issue, we give a reboot to server. But with the progress in time the frequency is getting increased drastically (approx once in 2- 3 days).
Any pointers on how to solve this issue will be greatly appreciated.
Thanks In Advance,
Vipin.
Hi,
First of all tune the poral parameter as per SAP note 723909 - Java VM settings for J2EE 6.40/7.0
Secondly tune the parameter as per belowlink
http://www.sdn.sap.com/irj/scn/index?rid=/library/uuid/f0f1358d-0812-2c10-b58c-c7bdd7a0cdce
Thanks
Anil
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi vipin,
There may many things for this i.e Sizing or parameter may not be tunned properly.
I don't thing there could be sizing problem, becuase the server detail which you have provided says that sizing of server is good.
Can you tell me did you created any additional server node or you are using only 1 server node.
If your are using only 1 server node then it is better to add 2 more server node and tune the parameter of all server node as per the note provided earlier.
Definetly it will resolve your performace problem.
Thanks
Anil
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Anil, Juan,
Thanks for your valuable inputs!
As per Anil' s suggesstion, i planned to add extra server nodes to the server but unfortunately over the weekend server went down again and had to give a reboot to bring it back up and running fine. Now this has raised doubts in my client' s mind that if it was an issue with respect to load then over the weekend the number of users who access portal are very few and statting it might be some other reason than load.
Now the issue seems to end up by raising an OSS notes with SAP and i have been given responsibility to prepare a doc that needs to be submited to SAP.
I have already got the screenshots from SAPMMC on it' s status and all, but also requires log information specific to the reason behind this behaviour. Can i know where the log information specific to such issues gets logged and all the necessary information that i need to document to raise an OSS.
Inputs with be of great help and will be highly appreciated.
Thanks In Advance,
Vipin.
Hi Vipin,
Well, you issue certainly related with sizing the environment.
And also you can check how you GC is working.
But to report your issue, i strongly suggest WAIT for issue to re-appear then
Goto - SAPMMC
Expand - J2EE Prcess Table
Select the 'ServerX' -> Choose Actions (menu option)-> Select 'Dump Stack Trace'
This exactly captures the issue and helps SAP to analize further.
Along with this, you may also required to send 'Defaulttrace.trc' file as well as its tracks all happenning things in the envrionment.
and enclose even, System Information (Component Version information) this you can get from the default page.
I hope this helps you.
Regards
Sekhar
To drill down to the exact cause of the issue, u better install some 3rd party tools, supported by SAP, like Wilyintroscope. With Willyintroscope, you can see the response time at each step, when a request is placed. Also, thru 'jpmon', you can get detail status regarding the thread servers. <br /> <br /> when ever you see server status in 'red', try to print the thread dump, using jpmon and try restarting only that individual thread server. <br /> <br /> Also, there are many SAP Notes, which suggest to go for latest SP & kernel, available in marketplace. With which, most of the know issues will be fixed (likes thread getting stalled). <br /> <br /> check note:1016241, 885645 , 599539 & 1146711<br /> <br />
Hope this info helps you to fix the issue.<br /> with regards,<br /> Rajesh,.<br />
Edited by: Rajesh Mutharaju on Feb 10, 2010 1:16 PM
Hi all,
Thanks for all your helpful responses! well, the issue has been tracked, it was because of the reason that the oracle listener TLSLSNR which was going down, users were unable to access portal via browser and a way http provider componenet inside SAPMMC was going into red status.
For now DBA team are increasing the number of processes from their side to resolve the issue.
Thanks Once Again,
Vipin.
User | Count |
---|---|
83 | |
24 | |
12 | |
9 | |
7 | |
6 | |
5 | |
5 | |
5 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.