Error while loading ML model in ElasticSearch

General Feedback
21

I have a clater with this nodes:
10.31.57.25 53 99 2 0.77 0.82 0.87 d data,ml - mn-insideelkens002
10.31.56.252 24 84 0 0.00 0.04 0.01 m master - mn-insideelkensmstr003
10.31.57.188 6 76 0 0.10 0.05 0.01 m master - mn-insideelkensmstr001
10.31.56.149 11 84 0 0.09 0.06 0.01 m master * mn-insideelkensmstr002
10.31.56.104 11 99 2 1.09 1.09 1.04 d data,ml - mn-insideelkens001
10.31.56.198 26 99 0 0.02 0.01 0.00 - - - mn-insideelkenscrd001

And cluster config is :

{
“persistent”: {
“plugins”: {
“ml_commons”: {
“only_run_on_ml_node”: “false”,
“sync_up_job_interval_in_seconds”: “5”,
“native_memory_threshold”: “100”
},
“index_state_management”: {
“metadata_migration”: {
“status”: “1”
},
“template_migration”: {
“control”: “-1”
}
}
}
},
“transient”: {},
“defaults”: {}

22

My claster is isolated from internet.
Log file:

1
11215×861 29.6 KB

23

Hey @TonyStark

I had a issue like yours today. ML model was not responding, In my case it was memory allocated, just ran out. But my quick fix was

POST /_plugins/_ml/models/lmIqgocBV8bdr9yTG83x/_load

it came back to life,

24

@Gsmitt Thanks for your help. @TonyStark Do you see more issues?

Some small suggestions, how about changing the title to “Error while loading ML model in OpenSearch” ?
BTW, we have a dedicated ML category Machine Learning - OpenSearch, will be best if can move this question there.

1 Like
25

Hey @TonyStark, have you ever solved the Connection timed out issue?