Version: Elasticsearch 6.8
I have a few indices that cannot be frozen, the freeze operation returns with:
{
"acknowledged" : false,
"shards_acknowledged" : false
}
Inspecting the index yields the following interesting property:
GET myindex/_settings
...
{
"myindex" : {
"settings" : {
"index" : {
...
"frozen" : "false",
...
}}}
I have been using the freeze feature for a while now, but I'm pretty much out of clues how to proceed with this phenomenon.... Maybe my scheduled job borked these indices... a cluster event interrupted the freeze process?
Things that I checked/tried:
- checking the node vs routign allocation requirement -> OK
- checking if the index is read-only -> it's not RO
- querying the index as a live index -> works
- closing the index -> acknowledged: false (i.e. doesn't work)
- freezing the index -> acknowledged: false (i.e. doesn't work)
- checking the deug level logs -> no useful information
- opening the index -> acknowledged:true, but makes no difference (and close/freeze won't start working after this)
Thanks for any help/idea you may offer.