Details
-
Bug
-
Resolution: Fixed
-
Major
-
Cheshire-Cat
-
Enterprise Edition 7.0.0 build 2497 ‧ IPv4 © 2020 Couchbase, Inc.
-
Untriaged
-
Centos 64-bit
-
-
1
-
No
Description
Script to Repro
./testrunner -i /tmp/win10-bucket-ops.ini rerun=False,crash_warning=True,quota_percent=95 -t bucket_collections.collections_rebalance.CollectionsRebalance.test_data_load_collections_with_rebalance_in_out,nodes_init=4,nodes_in=2,nodes_out=1,bucket_spec=multi_bucket.buckets_for_rebalance_tests_more_collections,data_load_spec=volume_test_load_with_CRUD_on_collections,data_load_stage=before,quota_percent=80,GROUP=rebalance_with_collection_crud
|
Basically we add 2 nodes and remove 1 node and do a rebalance. It fails with the crash with following bt. Interestingly even though bt has full of DurabilityMonitor, this test was run with non durable writes.
bt full
bt_full.txt
On a side note, also saw a few other interesting messages like the ones shown below.
[root@s71920 logs]# grep 'segment delete failed error:IOError' *
|
babysitter.log:memcached<0.115.0>: 2020-07-01T09:48:16.323573-07:00 CRITICAL WAL::Trim 18:6070272 segment delete failed error:IOError: Error (2) removing segment /opt/couchbase/var/lib/couchbase/data/bucket1/magma.3/wal/wal.17 - No such file or directory
|
memcached.log.000269.txt:2020-07-01T09:48:14.375924-07:00 CRITICAL WAL::Trim 18:6676480 segment delete failed error:IOError: Error (2) removing segment /opt/couchbase/var/lib/couchbase/data/bucket1/magma.0/wal/wal.17 - No such file or directory
|
memcached.log.000269.txt:2020-07-01T09:48:16.323573-07:00 CRITICAL WAL::Trim 18:6070272 segment delete failed error:IOError: Error (2) removing segment /opt/couchbase/var/lib/couchbase/data/bucket1/magma.3/wal/wal.17 - No such file or directory
|
grep: rebalance: Is a directory
|
[root@s71920 logs]#
|
cbcollect_info attached.