Did you restart the informatica services after the upgrade was done?
From which version did you upgraded to 9.5.1?
Yes, we have. it was running fine since 3 days. from yesterday night we got this waiting issues. Later after the integration service was restarted, all the jobs came to running state.
from 8.6.1 we upgraded to 9.5.1.It has been installed on new boxes 64-bit linux,64 GB Ram, 12 GB Swap, two 8-bit CPU servers. Repository was backed up and upgraded to new database.
Any idea, recommended RAM size for 64-bit installation ?
Check for the folloowing document to get information about the windows requirement for Informatica upgrade from 8.6 to 9.5
From the document it can be seen that 8GB of RAM is enough.. and since you have 64GB of RAM, i dont think configuration can be an issue..
Did you check your server CPU usage, when your sessions are going to waiting state? is it showing 100% of CPU usage?
And how many workflows are you executing at one time?
Are there any other processes running except the Inforamtica Power center on that box?
My installation is on linux. CPU usage is under control.
Around 20 workflows on one node, 9 workflows on that another node.. on an average around 10 run each on a node.
no, there are no other processes than informatica. we have java installed for few of our scripts using java classes.
are you facing this issue for all the workflows that you are running? or for any specific session?
Can you attach the session and workflow log of the session that is going in the waiting state..
May be it can be of some help..
can you please increase Maximum processes value at node level - properties tab. and see if waiting counts get decreased. and if still facing waiting then plz manage your workflows run and not to run many workflows parallely.
Nitin - with the same load, we ran fine for about four years in 8.6 environment. that too , old one has only 16 GB RAM with 8 GB Swap.
Logs says as below:
Request [:Session task instance [wrklt_SA_admin.s_m_SA_MVP_AM_CJMIOKLN_ADMIN]:] rejected  nodes [AL_Node1595 ] due to insufficient memory.
Can you check if in the mapping which is associated with this session s_m_SA_MVP_AM_CJMIOKLN_ADMIN, there are any ports that have very high precision..
Please confirm the nofile descriptor value currently configured on your Unix server where 951 is running. You can run the following command to find out the value.
Please read the KB 15597 to understand the resource limits on UNIX System
Also run the CSM for 951 environment. Follow the KB 32194 to generate the CSM.