- 
    Type:Bug 
- 
    Resolution: Done
- 
    Priority:Major - P3 
- 
    Affects Version/s: None
- 
    Component/s: WiredTiger
- 
        Fully Compatible
- 
        
- 
        Build C (11/20/15)
- 
        None
- 
        None
- 
        None
- 
        None
- 
        None
- 
        None
- 
        None
- 12 cores, 24 cpus, 64 GB memory
- 1-member replica set
- oplog size 5 GB, default cache
- data on ssd, -journal on separate hdd- no journal
- 5 threads inserting 1 KB random docs as fast as possible (see code below)
- recent nightly build:
2015-11-19T08:41:38.422-0500 I CONTROL [initandlisten] db version v3.2.0-rc2-211-gbd58ea2 2015-11-19T08:41:38.422-0500 I CONTROL [initandlisten] git version: bd58ea2ba5d17b960981990bb97cab133d7e90ed 

- periods where oplog size exceeded configured size start at B and D
- oplog size recovered during stall that started at C, but did not recover after D
- growth period starting at D appears to coincide with cache-full condition, but excess size starting at B looks like it may be simply related to rate of inserts
- is depended on by
- 
                    SERVER-21442 WiredTiger changes for MongoDB 3.0.8 -         
- Closed
 
-         
- 
                    WT-1973 MongoDB changes for WiredTiger 2.7.0 -         
- Closed
 
-         
- 
                    SERVER-21730 WiredTiger changes for 3.2.0-rc6 -         
- Closed
 
-         





