Details
-
Bug
-
Resolution: Fixed
-
Major
-
7.2.0
-
7.2.0-5227
-
Untriaged
-
Centos 64-bit
-
-
0
-
Unknown
Description
Steps To Recreate:
- Create a 3 node cluster
- Create a magma buckets with (bucket_history_retention_seconds=1500,bucket_history_retention_bytes=99000000000,) (vbuckets = 16, replicas = 2)
- Create 14 collections(total collection count is 15, including default collection)
- After creating collection update the collection history setting to true
- Create 5 million docs in each of the collection
- Upsert all the docs
- Again start upserting all the docs
- While doc ops are going on reduce the replicas to 1 and change fragmentation to 30% trigger rebalance @ 6:11:36 AM
- Again change the replicas to 2 and trigger rebalannce(6:13:20 AM)
- Immediately after above step tried magma_dump command and observed segmentation fault(@ 06:14:30 2023 from dmesg)
(Tried on node 172.23.107.76) (Would be helpful if we can have minidump facility for magma dump, I tried attaching gdb after this but no luck in reproducing this)/opt/couchbase/bin/magma_dump /data/default/magma.0 --kvstore 0 --docs-by-seq
Segmentation fault
demsg:
[Sun Mar 5 06:14:30 2023] magma_dump[40182]: segfault at 0 ip 0000000000466fa2 sp 00007fff9fba3010 error 4 in magma_dump[400000+3c1000]
|