Luminis crashing....
Thu, 2008-03-06 15:21 — Renata
We went live on Monday and had registration start on Wednesday. We have had Luminis "freeze" or "hang up" at least twice a day. We have to restart all the services then everything is fine. We have 16 gig ram on our Windows server. We notice that once it starts to use a little over 2 gig the system freezes. However our cpu is only using 1%. We would like to know where to start with this problem.
Thank you again.
Hack Your Knowledge
Recent blog posts
- LDI Queue Monitoring Script
- U of S Single Sign-on SSO Monitoring Script
- Lum IV - Perf problems and Paralled Deployment
- WebCT custom SSO Icon
- Examples using our AjaxRequests channel
- User-defined role and event trigger 101
- Silly question about lumdev itself
- See the XML nested-tables.xsl is processing
- Need Luminus version 3 or 4 for Installation
- Guest Login to IV?
Developer Links
Poll
What is your estimated peak concurrent usage (sessions) on Luminis at your institution?
More than 2500
13%
Between 2000-2499
0%
Between 1500-1999
0%
Between 1200-1499
0%
Between 900-1199
13%
Between 500-899
13%
Fewer than 500
63%
Total votes: 8
- Login or register to post comments
- Older polls
Syndicate
Memory Usage
What version of Luminis? Are you running parallel deployment or not?
With Luminis III, it doesn't matter how much memory you have at all, you can only use the first 4 GB for the JVM that Luminis runs in. I'm unsure whether Luminis IV has this same restriction.
Anyway, you need to partition the memory properly. Check your jvm12.conf in the web server config directory (something like /opt/pipeline/products/ws/https-*/config/ ). Here are the values we use in production:
jvm.maxHeapSize=2936012800
jvm.minHeapSize=2936012800
jvm.option=-XX:MaxNewSize=411041792
There are MANY other things that you need to take in consideration (ie: default channels in home tab, session timeout value... the list is long) but start with the above and proceed from there.
Todd