Jeff Johnson

I've looked at zfs/cmd/zed/agents/zfs_retire.c and I cannot identify why the resilver to hot-spare doesn't execute. Detailed zpool events: https://pastebin.com/13r4UDL6

Would the `zpool offline -f` generate the event message that zed agents are looking for? In cmd/zed/agents/zfs_diagnosis.c it appears to be looking for io, checksum or probe failures. I was...

There were SCSI errors (hardware/media on writes) ``` Jun 28 00:02:17 myhost kernel: sd 0:0:124:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE Jun 28 00:02:17 myhost kernel: sd 0:0:124:0: [sda] tag#0...

Here is a [pastebin](https://pastebin.com/6eebRTAa) of the `zpool events -v` I would think that ZED would've seen what it needed and engaged the S1 hot-spare after on the statechange at 00:02:19.107546574....

*poke* This still appears to be a defect, others reporting mixed results. We are seeing it at customer sites as well currently. Any thoughts on what/where to dig on this?

Uncommenting? Or commenting it out? Builds failed for me on 2.15.2 unless I commented out 854->858.