ThomasAlxDmy

Thanks for helping @yanliang567 @xiaofan-luan So I started to spin up a smaller environment so we can iterate faster and understand what's going on: Collection Loaded => 1,265,800 records. I...

Here's the etcd backup! (unfortunately cluster is ansible/terraform I haven't set up monitoring yet). [bw_etcd_ALL.230118-181249.bak.gz](https://github.com/milvus-io/milvus/files/10452895/bw_etcd_ALL.230118-181249.bak.gz)

https://github.com/milvus-io/milvus/blob/master/internal/datanode/flush_manager.go#L851-L852 @congqixia you might know about this, any idea of what the problem is and how it can be mitigated?

sounds about right: Here's the logs from data-coordinator node: ``` [2023/01/05 03:33:34.390 +00:00] [INFO] [datacoord/util.go:158] ["no vector index for the segment"] [collectionID=438376139142926320] [segmentID=438376139143574409] [2023/01/05 03:33:34.390 +00:00] [INFO] [datacoord/util.go:158] ["no vector...

Confirmed it's fixed!

+1 @bpot

https://github.com/FiloSottile/gvt/pull/79

sure thing, we could do that! PR updated :)

@FiloSottile ping. Is it good to go?

@FiloSottile Can you merge this please?