We are running into memory issues in one of our environment. When we stop and re

sbalasani over 12 years ago

We are running into memory issues in one of our environment. When we stop and restart Appian engines (including jboss) all the engines are running fine with Okay status but after few hours Appian is generating huge amounts of data in just a matter of few minutes and eventually running out of space and the system is crashing. I have checked the log file to see whats causing the issue but I only see errors which are related to the engines being down. Anyone had this kind of issue before? If so please let us know how to debug. Thanks.
We are using Appian 6.6.0....

Discussion posts and replies are publicly visible

0 Myles Weber
Appian Employee
over 12 years ago

I'd start with patching up. 6.6.1 has been out for 18 months and is required.
forum.appian.com/.../Appian_6.6.1_Release_Notes
forum.appian.com/.../Appian_6.6.1_Hotfixes
While at it, you are likely running Java 6 too, which doesn't get fixes unless you are paying Oracle for it.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 Owen Parrish
Appian Employee
over 12 years ago

You can use the sizing script to identify with processes are the largest. forum.appian.com/.../Identifying_Process_Memory_Usage
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 jorgea272 over 12 years ago

What are your hardware specs on the application server?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel
0 aneeshv over 12 years ago

yes, stop jboss. ..stop engines... you have to clean your kdbs. clear all kdbs and leave the latest one for each gateway or k process. make sure you do a checkpoint. Then restart engines/gws and then restart jboss.

Also, always perform sizing and delete process instances completed according to sizing estimates...

your process design should be such that you should have very little load in your main process, bring all required business data ( only that data to be shown in reports) to the main process, delete all sub-processes after completion and put all data for long term reports into an external db.

you should regularly run cleanup scripts for data(kdbs), archived processes and logs cleanup.. it should be a cron job on the server.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel