Details
-
Bug
-
Resolution: Fixed
-
Critical
-
5.1.0
-
Untriaged
-
Yes
Description
centos longevity - 5.1.0-5527 - test has been running for 18 hours - we use pillowfight and gideon to load data - usually we hit around 26M items on a 7 day run - default bucket is already showing 73,786,976,294,833,370,000 items in UI which is impossible to be loaded - checked bucket stats using cbstats tool:
[root@localhost bin]# ./cbstats 172.23.108.103:11210 all -u Administrator -p password | grep curr
|
curr_connections: 311
|
curr_conns_on_port_11207: 2
|
curr_conns_on_port_11209: 143
|
curr_conns_on_port_11210: 166
|
curr_items: 18446744073707788097
|
curr_items_tot: 18446744073707915516
|
curr_temp_items: 2
|
vb_active_curr_items: 18446744073707788097
|
vb_pending_curr_items: 0
|
vb_replica_curr_items: 127419
|
curr_items is showing 2^64 - not sure if this is a regression - seeing it for the first time in longevity though - attaching screenshots and logs - issue not seen in vulcan so far
Attachments
Issue Links
- relates to
-
MB-26567 Change ep-engine full eviction item counts to not overcount
-
- Resolved
-