compaction.redzone |
int |
To create a new redzone, write a percent value into compaction namespace. |
compaction.redzone.mode |
int |
Set compaction mode. Mode 1: branch-less mode (strict 2-level storage), 2: checkpoint, 3: branch + compaction (default). |
compaction.redzone.compact_wm |
int |
Compaction operation starts when a number of node branches reaches this watermark. Cant be less than two. |
compaction.redzone.compact_mode |
int |
Set read-intensive or write-intensive compaction strategy mode. 0 - by number of branches, 1 - by temperature. |
compaction.redzone.branch_prio |
int |
Priority of branch operation. Priority is measured by a maximum number of executing workers. |
compaction.redzone.branch_wm |
int |
Branch operation starts when a size of in-memory key index reaches this watermark value. Measured in bytes. |
compaction.redzone.branch_age |
int |
Branch operation automatically starts when it detects that a node in-memory key index has not been updated in a branch_age number of seconds. |
compaction.redzone.branch_age_period |
int |
Scheduler starts scanning for aged node in-memory index every branch_age_period seconds. |
compaction.redzone.branch_age_wm |
int |
This watermark value defines lower bound of in-memory index key size which is being checked during branch_age operation. Measured in bytes. |
compaction.redzone.anticache_period |
int |
Check for anti-cache node election every anticache_period seconds. |
compaction.redzone.backup_prio |
int |
Priority of backup operation. Priority is measured by a maximum number of executing workers. |
compaction.redzone.gc_wm |
int |
Garbage collection starts when watermark value reaches a certain percent of duplicates. When this value reaches a compaction, operation is scheduled. |
compaction.redzone.gc_db_prio |
int |
Priority of a database async close/drop operation. |
compaction.redzone.gc_prio |
int |
Priority of gc operation. Priority is measured by a maximum number of executing workers. |
compaction.redzone.gc_period |
int |
Check for a gc every gc_period seconds. |
compaction.redzone.lru_prio |
int |
Priority of LRU operation. Priority is measured by a maximum number of executing workers. |
compaction.redzone.lru_period |
int |
Run LRU scheduler every lru_period seconds. |
compaction.redzone.async |
int |
Asynchronous thread work mode: 1 - reserve thread, 2 - do not reserve thread. |