We have installed Datastage server 7.5.1A on solaris 9 OS .Our Development evironment has V490 server machine and production has E2900 server machine.
Whether the kernel parameter are dependent on the server model besides the OS.
Do we need to add any other kernel parameter on E2900.Currently following kernel parameters are present on E2900 (solaris 9)
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=18000
set semsys:seminfo_semmsl=3400
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmni=2000
set shmsys:shminfo_shmseg = 200
set shmsys:share_page_table=0
set noexec_user_stack=1
*
set msgsys:msginfo_msgmax = 8192
set msgsys:msginfo_msgmnb = 16384
set maxuproc = 400
Thanks
Kernel parameters
Moderators: chulett, rschirm, roy
-
- Participant
- Posts: 54607
- Joined: Wed Oct 23, 2002 10:52 pm
- Location: Sydney, Australia
- Contact:
Provided the numbers are at least as large as those in the documentation you received with the software, and provided that there are no other significant consumers of shared memory, semaphore or message queues (that is, DataStage server is the principal application on the machine), then the numbers ought be OK.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
[quote="ray.wurlod"]Provided the numbers are at least as large as those in the documentation you received with the software, and provided that there are no other significant consumers of shared memory, semaphore
The parameter values are as per the minimum requirements.We have installed oracle serrveer and datasstage on the same box.So how can we calculate the correct value for the parameter considering this..
Please suggest.Is there any document by Datastage?
The parameter values are as per the minimum requirements.We have installed oracle serrveer and datasstage on the same box.So how can we calculate the correct value for the parameter considering this..
Please suggest.Is there any document by Datastage?
[quote="ArndW"]psv,
if you installed according to the minimum kernel settings specified in the DataStage upgrade and installation guide on page 41 then what exactly is your question? I would not necessarily go ab ...[/quote]
My question is that if the server model changes ( V490 to E2900)and OS is same then , is it going to affect the parameter values..
E2900 is a high end server then v490.
If other application processes are utilizing kernel parameters defining for Datastage ,then how to calculate actual value of kernel parameter.
I hope the question is clear.
Thanks
if you installed according to the minimum kernel settings specified in the DataStage upgrade and installation guide on page 41 then what exactly is your question? I would not necessarily go ab ...[/quote]
My question is that if the server model changes ( V490 to E2900)and OS is same then , is it going to affect the parameter values..
E2900 is a high end server then v490.
If other application processes are utilizing kernel parameters defining for Datastage ,then how to calculate actual value of kernel parameter.
I hope the question is clear.
Thanks
In this case I would keep the kernel parameters the same. If you find that after a period of measurement that your system is not being fully utilized you can think about which kernel parameters (if any) might be changed to make DS work more efficiently in the new hardware environment. But I would not recommend making any initial changes as long as everything was working on the old platform and still works on the new.
-
- Participant
- Posts: 407
- Joined: Mon Jun 27, 2005 8:54 am
- Location: Walker, Michigan
- Contact:
maxuproc is used to control how many concurrent processes a login can execute. I believe that 100 is way to low for most environments. I think that 500 would be a more reasonable value.
The value you choose is going to be dependant upon how many nodes your jobs use and how many jobs will run concurrently. Just based upon the number of cpus you have, I don't think that 1200 is an unreasonable value.
The value you choose is going to be dependant upon how many nodes your jobs use and how many jobs will run concurrently. Just based upon the number of cpus you have, I don't think that 1200 is an unreasonable value.