RHEL eating RAM space in production environment (top command)
Dear Members,
I have attached the top command output and free command output with the message. I see all RAM is being consumed. What can be the issue for this? would be glad for your help. |
|
Quote:
Quote:
|
I believe one has to have more than 1 post before they can attach files...
That said, if this a RHEL server, have you contacted RedHat Support? |
Quote:
|
Quote:
|
Using RHEL 5.7.
The output of top command (SHIFT+M) is as follows: Code:
top - 08:40:17 up 3 days, 20:26, 1 user, load average: 18.22, 18.67, 18.38 # free -m command gives the following output: Code:
total used free shared buffers cached |
Please use code tags. See my signature if you don't know how.
Code:
08:40:17 up 3 days, 20:26, 1 user, load average: 18.22, 18.67, 18.38 |
Your CPU is currently 68% idle. Memory is practically depleted; almost nothing is available or used for the buffer cache. I conclude that your application uses up memory. You say
Quote:
I see 42 processes running, each of them using 1GB or a bit more RAM, totaling 45GB. I wonder how many PSAPPSRV are running altogether? Couldn't they add up to 128GB? Hint: you can scroll top with pgup/pgdown, or run ps -el | grep PSAPPSRV. |
What about "uname -a".
And yes, why all the (PeopleSoft ?) instances if only for a few users ?. |
Quote:
|
Quote:
EDIT: I misread the free output. It makes sense now. Regarding your question, the kernel doesn't just launch PSAPPSRV processes because it feels like it. This is most definitely an application issue. It could be caused by a condition outside of the application, but still, the application should be able to handle external conditions correctly and not start tens of processes if they are not needed. EDIT: I just saw that PSAPPSRV is probably a Peoplesoft process. I would ask the question in an Oracle or Peoplesoft forum. |
Quote:
|
Quote:
|
just a question:
is this only one app with several threads or a lot of apps? (see -H in top) |
All times are GMT -5. The time now is 08:36 AM. |