2

I have migrated my existing data in 4 nodes Cassandra (with RF=3) to Elassandra and after putting my mappings whole data got indexed into Elassandra. After the completion of indexing, all nodes show a consistent result in /_cat/indices?v API. But as soon as I restart any node the data on that node is reduced substantially, index size as well as the number of records. If I restart another node of the cluster the problem shift to that node and previous node recovers automatically. For more details and detailed use case please refer to the issue I have created with Elassandra.

Pankaj Yadav
  • 139
  • 1
  • 10
  • Please share the index settings, especially the shard and replica count – ibexit Feb 20 '20 at 12:01
  • No. of shards=4, replica count=2. I have shared the full cluster details in the issue I logged with Elassandra. I was able to reproduce this on 10 shards, 2 replica setup as well. The concerned index name is cmsentitydb. Please check the file attached https://github.com/strapdata/elassandra/files/4226477/cluster-state.txt – Pankaj Yadav Feb 20 '20 at 17:36

1 Answers1

0

Upgrade to Elassandra v6.8.4.3 has resolved the problem. Thanks!

Pankaj Yadav
  • 139
  • 1
  • 10