-
Enhancement
-
Resolution: Fixed
-
P4
-
14
-
b18
Better concurrent refinement statistics are needed in order to improve the control of the concurrent refinement threads and the mutator thread processing of dirty card buffers. We need estimated card logging rates rates and estimated logged card concurrent processing rates.
Currently the nearest we have is DCQS::processed_buffers_mut, which is not at all sufficient. And once we have sufficient information, that buffer count should be replaced.
Currently the nearest we have is DCQS::processed_buffers_mut, which is not at all sufficient. And once we have sufficient information, that buffer count should be replaced.
- relates to
-
JDK-8234208 Logging reports zero total refined cards under "Before GC RS summary"
- Resolved
-
JDK-8232232 G1RemSetSummary::_rs_threads_vtimes is not initialized to zero
- Resolved
-
JDK-8137022 Concurrent refinement thread adjustment and (de-)activation suboptimal
- Resolved
-
JDK-8224169 Cleanup G1DirtyCardQueueSet::num_processed_xxx
- Closed