Hi all I've a test cluster with 6 nodes with 80GB of PVC each. And i've the following behaviour which is very strange :
I've one shard on a a node this shard is about 15GB and the free size is about 18GB so I've 47GB which have disappeared.
Any help would be appreciated for understanding why I've lost this space.
Nope. But today i've an issue with my ECK and it seems that the huawei storage system which is the under layer of the storageclass system seems to have some trouble. After some actions it seems to go back correctly with the right space available. Don't know exactly why ... but need to monitor that correctly.
Waiting my new storage system with block mode access (iSCSI, NVMe-TCP) for having a correct access and the needed performance.
I need to configure the right balancing on my ECK Actually on the 6 hot nodes I've about 30/35 shards on 3 nodes and the other 3 have only 2 shards ... dont't know exactly why ...
I also need to open a noew topic on ILM ... which is linked to storage due to consumption on shards
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.