Versions : 3.7.0
Describe the issue :
Hello folks!
Run local cluster using the sample docker-compose from docs Docker | OpenSearch Documentation
Load data that exceeds the configured memory limit, for example, а 1Gb index
Start \_update_by_query operation from Dev Tools
The operation returns plenty of version_conflict_engine_exception errors. If the index size is smaller, the query runs just fine.
Questions:
Why does this happen?
What process modifies the index during update? I’m running only a single request on a local cluster and doing nothing else.
How can I debug this?
pablo
August 4, 2026, 11:41am
2
Hi @Smoking_Machine , welcome to the forum!
Could you share your full _update_by_query query?
Hello!
My query looks like
POST test-idx-1/_update_by_query
{
"query": {
"term": {
"groups.billing_group.keyword": "forwarding"
}
},
"script": {
"source": "ctx._source.groups.billing_group = params.grp",
"lang": "painless",
"params": {
"grp": "forwarding2"
}
}
}
pablo
August 6, 2026, 1:09pm
4
@Smoking_Machine Did you consider using sroll_size and slices?
I’ve tested with a 3 GB index and 46K documents and I’ve got all documents updated without version conflicts.
i.e.
POST /test-logs/_update_by_query?slices=10&scroll_size=100&conflicts=proceed&refresh=true
{
"query": {
"match_all": {}
},
"script": {
"source": "ctx._source.test_update_counter = 5"
}
}