Derek Su

@sushant-suse Please study the hardened cluster background knowledge and setup. Draft a hardening guide like https://github.com/harvester/harvester/issues/8792. Thanks.

@opipenbe Thanks for the request. Just want to gather more information in your use case. Does the engine-replica-timeout parameter work well if the storage path failovers can be completed with...

> Hi, yes [@derekbit](https://github.com/derekbit) , we have seen more stability while increasing engine-replica-timeout value from default to 30-seconds. Although this time I observed that 30-seconds was not enough. Thanks @opipenbe....

Added `require/auto-e2e-test` label. It might need to backport v1.8.x and v1.7.x. Let's check the scope of the improvement later.

> Hi [@ChanYiLin](https://github.com/ChanYiLin) > > While this situation occurs, the correct workaround is to delete related bids? Thanks cc @COLDTURNIP

@COLDTURNIP Can you provide ta workaround for the case firtst? We can reprioritize the ticket if there is a valid workaround. Thanks.

cc @houhoucoop @a110605 @innobead

Found lots of error message ``` 2024-05-17T00:23:51.008868782Z [pvc-a5eee114-96e6-4235-87ec-b9ae73e9cc85-r-27d66cce] time="2024-05-17T00:23:51Z" level=error msg="Failed to read" func="dataconn.(*Server).readFromWire" file="server.go:42" error="wrong API version received: 0x5250" 2024-05-17T00:23:54.460279343Z [pvc-a5eee114-96e6-4235-87ec-b9ae73e9cc85-r-27d66cce] time="2024-05-17T00:23:54Z" level=info msg="New connection from: 10.42.2.206:59388" func="rpc.(*DataServer).listenAndServeTCP" file="dataserver.go:57"...

@ejweber Could you help check the issue? It might be related to https://github.com/longhorn/longhorn/issues/8289 that Shou found it in a user's environment. Thank you. cc @shuo-wu @PhanLe1010 @innobead

> @derekbit and @innobead, I think we should investigate this more to try to make improvements. However, there doesn't seem to be any evidence of a regression here in my...