Whenever I try to perform a shrink operation on an index using ISM I always get the error " There are no available nodes to move to to execute a shrink. Delaying until node becomes available." I’ve 2x-checked my nodes and they have way more than ample space to hold the 2x-shard size required to perform a shrink.
I’ve tried this on a non-monolithic cluster (dedicated data nodes, coord-nodes, master-node, warm-nodes) and also on a simple 4-node cluster where all nodes are performing all of the duties and I’m getting the same error.
I’ve looked through the kotlin code and all logic seems correct and I can’t find any glaring logic bugs. I’ve also looked through the doco to make sure that all prerequisites are being met. What am I doing wrong? Has anyone else gotten the shrink ISM operation to work correctly?
I am curious if anyone else has been able to get the Shrink operation working correctly. And if so, how they set it up.
I believe I’ve discovered a bug in the ISM Shrink action code that is causing this problem. I created a bug report here: https://github.com/opensearch-project/index-management/issues/417 . I’m not a Kotlin dev so I’m having problems getting the integration tests to complete. Would welcome anyone’s help if they can so I can make a PR/MR to get this fixed.
anyone managed to get it working? I just updated to Opensearch 2.2.0 and I’m still getting the same error.
I deployed it on a Kubernetes cluster with 3 data nodes and 30GB PVC for the data mountpoints
$ curl localhost:9200/_cat/allocation?v
shards disk.indices disk.used disk.avail disk.total disk.percent host ip node
79 469.9mb 733.5mb 28.6gb 29.4gb 2 10.39.23.27 10.39.23.27 opensearch-data-0
83 672.3mb 1006.6mb 28.4gb 29.4gb 3 10.39.17.161 10.39.17.161 opensearch-data-2
33 603.2mb 778.3mb 28.6gb 29.4gb 2 10.39.26.25 10.39.26.25 opensearch-data-1
verified on 2.2.0 - shrink does work.
can you show me how does your policy look, please? maybe I’m doing something wrong