JMast
Reputable Poster
Hello All,
We upgraded to 9.2 last winter. We noticed that occasionally JDE was crashing. I discovered that the jvm would be running out of heap space in a garbage collection panic situation. I don't know anything about weblogic garbage collection, so I had to get educated quickly. I found that our servers were set with what looks like default xms32m xmx1024m settings on startup. I changed those to xms8196m and xmx8196m. However, over the course of time, the percent free space still declines each day. We are a one shift business Mon-Fri, so there should be plenty of time for threads from JDE to die and go inactive. It seems to me this must mean there are threads that are still considered live and not gc'd. Has anyone else seen this or have other ideas as to why the jvm would not be cleaning up?
One strange thing is that no matter what gc logging parameters I try to add to startup, the webconsole just rejects them. Any help on proper syntax or settings to log gc would be great.
We are on windows 2012, weblogic 12.2.1.0.0 java hotspot 25.161 (1.8.0.161
Thanks for any advice, if I need to provide more info, let me know.
Jer
date server heap size free current free percent
4/26/2018 8:02 c 8003256320 6271285312 78
4/26/2018 8:02 b 8014266368 7203672232 89
4/26/2018 8:02 a 8088715264 7027940096 86
4/26/2018 10:37 c 8047820800 6548115432 81
4/26/2018 10:37 b 8023703552 6315795352 78
4/26/2018 10:37 a 8099201024 6595865416 81
4/26/2018 16:27 c 8157396992 6636583272 81
4/26/2018 16:27 b 8066170880 6701362896 83
4/26/2018 16:27 a 8122793984 5177579936 63
4/27/2018 17:01 c 8200388608 5477935328 66
4/27/2018 17:01 b 8060928000 5338975680 66
4/27/2018 17:01 a 8144814080 5063097400 62
4/30/2018 10:14 c 8147435520 4444355280 54
4/30/2018 10:14 b 8086618112 3619803368 44
4/30/2018 10:14 a 8143241216 4357423712 53
We upgraded to 9.2 last winter. We noticed that occasionally JDE was crashing. I discovered that the jvm would be running out of heap space in a garbage collection panic situation. I don't know anything about weblogic garbage collection, so I had to get educated quickly. I found that our servers were set with what looks like default xms32m xmx1024m settings on startup. I changed those to xms8196m and xmx8196m. However, over the course of time, the percent free space still declines each day. We are a one shift business Mon-Fri, so there should be plenty of time for threads from JDE to die and go inactive. It seems to me this must mean there are threads that are still considered live and not gc'd. Has anyone else seen this or have other ideas as to why the jvm would not be cleaning up?
One strange thing is that no matter what gc logging parameters I try to add to startup, the webconsole just rejects them. Any help on proper syntax or settings to log gc would be great.
We are on windows 2012, weblogic 12.2.1.0.0 java hotspot 25.161 (1.8.0.161
Thanks for any advice, if I need to provide more info, let me know.
Jer
date server heap size free current free percent
4/26/2018 8:02 c 8003256320 6271285312 78
4/26/2018 8:02 b 8014266368 7203672232 89
4/26/2018 8:02 a 8088715264 7027940096 86
4/26/2018 10:37 c 8047820800 6548115432 81
4/26/2018 10:37 b 8023703552 6315795352 78
4/26/2018 10:37 a 8099201024 6595865416 81
4/26/2018 16:27 c 8157396992 6636583272 81
4/26/2018 16:27 b 8066170880 6701362896 83
4/26/2018 16:27 a 8122793984 5177579936 63
4/27/2018 17:01 c 8200388608 5477935328 66
4/27/2018 17:01 b 8060928000 5338975680 66
4/27/2018 17:01 a 8144814080 5063097400 62
4/30/2018 10:14 c 8147435520 4444355280 54
4/30/2018 10:14 b 8086618112 3619803368 44
4/30/2018 10:14 a 8143241216 4357423712 53
Last edited: