You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
"The earliest JVM detects idle state is after it has been idle for around 1 minute (but it is subject to change and it is not exposed to users). So any value less than 1 minute passed to -XX:IdleTuningMinIdleWaitTime gets rounded to 1 minute. Also, because of the way OpenJ9 currently detects its idle state, it may take anywhere around 'n' to 'n+100' secs to detect idle state where 'n' is the value specified to XX:IdleTuningMinIdleWaitTime option."
"When OpenJ9 releases memory pages during idle state, it does not modify internal GC structures which keep track of committed pages. It just de-commits the memory pages to OS using madvise() call. That's the reason Committed value does not reduce while RSS usage goes down during idle state. "
The text was updated successfully, but these errors were encountered:
Subject:
https://www.eclipse.org/openj9/docs/xxidletuningminidlewaittime/
Suggestion:
From the slack channel:
"The earliest JVM detects idle state is after it has been idle for around 1 minute (but it is subject to change and it is not exposed to users). So any value less than 1 minute passed to -XX:IdleTuningMinIdleWaitTime gets rounded to 1 minute. Also, because of the way OpenJ9 currently detects its idle state, it may take anywhere around 'n' to 'n+100' secs to detect idle state where 'n' is the value specified to XX:IdleTuningMinIdleWaitTime option."
"When OpenJ9 releases memory pages during idle state, it does not modify internal GC structures which keep track of committed pages. It just de-commits the memory pages to OS using madvise() call. That's the reason Committed value does not reduce while RSS usage goes down during idle state. "
The text was updated successfully, but these errors were encountered: