Details
-
Bug
-
Resolution: Fixed
-
Critical
-
5.5.0
-
Untriaged
-
Unknown
-
CX Sprint 97, CX Sprint 98
Description
In a recent QE reported issue (MB-28887) where we are unable to determine the start of a problem due to log rollover, over 141MB out of 244 MB log are 4086 instances of:
2018-03-23T22:48:52.651-07:00 INFO CBAS.metadata.MaintainDcpCallbackFactory [Executor-240:8800eb0ca0e77c7859db50c99b6ea7da] The DCP state written to {"class":LSMBTreeWithBloomFilterDiskComponent", "index":{"class":"DiskBTree","file":"storage/partition_0/Default/ds4/0/ds4/2018-03-23-22-48-52-616_2018-03-23-22-48-52-616_b"}} is ...
|
This is way too much logging. We need to either reduce this to DEBUG, or figure out how to greatly reduce the size of these logs. Were we to log an array instead of map, that would help a tiny bit, but still it will be way too much to appear in the product log.
Attachments
Issue Links
- relates to
-
MB-27314 Configure out of box log level of Analytics to WARN
- Closed
For Gerrit Dashboard: MB-28904 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
91794,3 | MB-28904 Redirect DCP related logs to analytics_dcp.log | vulcan | asterix-opt | Status: ABANDONED | -2 | +1 |
91984,5 | MB-28904: move verbose annoying dcp state logs to analytics_debug.log | vulcan | asterix-opt | Status: MERGED | +2 | +1 |
91989,2 | MB-28904: supply log direction as system property to driver | vulcan | cbas | Status: MERGED | +2 | +1 |
91998,2 | MB-28904: retain analytics_debug.log in cbcollect_info | master | ns_server | Status: MERGED | +2 | +1 |
92062,2 | MB-28904: fix max files to 10 | vulcan | asterix-opt | Status: MERGED | +2 | +1 |