Page Cache getting full on server (Linux)

krishnak756 over 6 years ago

Hello,

We have Appian 17.4 running on RHEL 6.8 server with 32 GB RAM and 8 CPU. The app server is JBoss (eap 6.4) and JDK (1.8.0_121). As soon as the appian is up and running, the memory and cache is almost getting full (90%). Even if we add more memory, it's still going up to 90% and as users (less than 10) start using it, memory is going down to 95% or more.

My questions are:

- Can we run Linux "page cache" script on the server every few hours? If not, why? Would this create any issue for the application?

- Are there are best practice or recommendations for memory tuning on Appian

- Are there any logs to see for memory leaks

Discussion posts and replies are publicly visible

0 Eliot Gerson
Appian Employee
over 6 years ago

If you run `top` , what are the biggest consumers of memory, and how much memory are they using?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 krishnak756 over 6 years ago in reply to Eliot Gerson

Thank you for your response Eliot.

The biggest memory consumers are: Appian execution, analytics and collaboration (around 6 GB) and Jboss ( 4 GB). The Jboss parameter is set to out of the box (min 2 GB and max 4 GB). Should we update the max to higher number? If so, how would that help Appian processes?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 Eliot Gerson
Appian Employee
over 6 years ago in reply to krishnak756

Are you saying that those processes are 6GB combined, or 6GB each? (Also, if it's 6GB each, do you mean 6GB per execution and analytics engine, or 6GB total for all execution engines?) If it's 6GB combined, and JBoss is using 4GB, then that makes me think that 1) something other than Appian is using the other ~20GB or 2) you may have made a mistake in how you're calculating memory usage ( see community.appian.com/.../kb-1149-how-to-analyze-memory-usage-on-linux-systems )
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel