Details
-
Bug
-
Resolution: Fixed
-
Critical
-
6.6.2
-
6.6.2->9588 to 7.0.0-5033
-
Triaged
-
Centos 64-bit
-
1
-
No
Description
Steps to Repro
1. Run longevity system test on 6.6.2 for 3 days. It initially had 27 node cluster.
./sequoia -client 172.23.96.162:2375 -provider file:centos_third_cluster.yml -test tests/integration/test_allFeatures_madhatter_durability.yml -scope tests/integration/scope_Xattrs_Madhatter.yml -scale 3 -repeat 0 -log_level 0 -version 6.6.2-9588 -skip_setup=false -skip_test=false -skip_teardown=true -skip_cleanup=false -continue=false -collect_on_error=false -stop_on_error=false -duration=604800 -show_topology=true
|
2. Swap rebalance 6 nodes one of each service type.
3. Now failover 6 nodes, one of each service type.
4. sysemctl stop couchbas-server on 6 nodes.
5. Install 7.0.0 on all of them.
6. Do a recovery + rebalance
7. Now failover 4 nodes, one of each service type.(172.23.106.117 172.23.105.25 172.23.105.210 172.23.105.206)
8. Do a delta recovery(for kv) + rebalance.
The above steps were repeated few times and finally we did upgrade using regular rebalance, i.e rebalance in of all the 7.0 nodes followed by another rebalance where we remove all the 6.6.2 nodes.
Post upgrade on the following node saw -ve values for raw_data_size.
172.23.106.225 : index
stats : [(u'default:default_result_ratings_claims_pindex:raw_data_size', -276), (u'default:default_result_ratings_claims_pindex (replica 1):raw_data_size', -278), (u'default:default_result_ratings_claims_pindex (replica 1):backstore_raw_data_size', -139), (u'default:default_result_ratings_claims_pindex:backstore_raw_data_size', -138), (u'(65-256)', -1), (u'(65-256)', -1)]
|
cbcollect_info attached.