version 7.14
I have some indexes that are not rolling over when I expect them to:
policy says: ` "max_primary_shard_size": "1gb",
and the stats show Storage Size 1.2gb and Primaries 1. Perhaps significantly there is no value for Primary Storage Size
You also have a replica over there, so storage size would be both of them combined, the cat shards api should return some info about shard size usage.
You can also try triggering the rollover api manually with dry run and the same conditions as in your ilm policy, that should tell you which condition was false
I also noticed that if you have multiple primaries this primary index size in ilm uses the size of only one of them to trigger the rollover, that had me a bit confused as well.
DOH, of course. And I notice that the size limit is depreciated.
I will just use the size I want divided by number of replicas.
Rolling over indices that small is not really ideal, you should be aiming for much larger ones to make sure you are using heap efficiently.
Ah! OK, I was not aware of that. Thanks Mark!
What sort of size is reasonable? hmmm... google
https://www.elastic.co/guide/en/elasticsearch/reference/current/size-your-shards.html
OK I will set the defaults back to 10 GB
30-50GB is what we recommend.
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.
© 2020. All Rights Reserved - Elasticsearch
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant logo are trademarks of the Apache Software Foundation in the United States and/or other countries.
