SLOG faulted. Same replacement process as regular disk?

[cryptyk]

Here's what I have:
NAME STATE READ WRITE CKSUM
Jabba DEGRADED 0 0 0
mirror-0 ONLINE 0 0 0
gptid/f3ecede4-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
gptid/f4813f8b-6bed-11e6-985e-000c296e056b ONLINE 0 0 5
mirror-1 ONLINE 0 0 0
gptid/f51c4964-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
gptid/f5b036c1-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
mirror-2 ONLINE 0 0 0
gptid/f6470f2f-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
gptid/f6e3d186-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
mirror-3 ONLINE 0 0 0
gptid/4f8cc1aa-24b3-11e7-a0f7-000c296e056b ONLINE 0 0 0
gptid/f81888ab-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
mirror-4 ONLINE 0 0 0
gptid/f8b3cf1f-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
gptid/df5f5f29-a346-11e7-b49a-000c296e056b ONLINE 0 0 0
mirror-5 ONLINE 0 0 0
gptid/d6cc7c7e-533e-11e8-9b0b-000c296e056b ONLINE 0 0 0
gptid/fa7efa09-6bed-11e6-985e-000c296e056b ONLINE 0 0 0
logs
gptid/3a02af31-b913-11e6-823d-000c296e056b FAULTED 0 8 0 too many errors


This is a home server so it's not mission critical. Everything is fully backed up to the cloud. I don't remember much about the initial setup I put together a few years ago, but it appears that I put the ZIL on an SLOG. It's an HGST 16GB SSD specifically made to handle ZFS ZIL workloads. It looks like it's failing, which kind of sucks, considering it's really expensive.

My understanding is that when an SLOG fails, the ZIL falls back to the pool, so I should be ok for the moment.

I have a couple of questions:
1) Do I replace this SLOG the same way I would a regular drive failure? Just offline it and replace it and the pool will resilver, etc?
2) If I decide NOT to replace it, can I just offline/remove it and permanently move the ZIL back to the pool? Are there any additional steps?

Thanks so much!

 

[Scharbag]

You can either replace it like any other disk or simply detach it if you no longer want SLOG.

 

[Chris Moore]

Here's what I have:

When you have a list like this to post, pleas use CODE tags so the list is formatted like this:

Code:

NAME											STATE	 READ WRITE CKSUM
Jabba										   DEGRADED	 0	 0	 0
  mirror-0									  ONLINE	   0	 0	 0
	gptid/f3ecede4-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
	gptid/f4813f8b-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 5
  mirror-1									  ONLINE	   0	 0	 0
	gptid/f51c4964-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
	gptid/f5b036c1-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
  mirror-2									  ONLINE	   0	 0	 0
	gptid/f6470f2f-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
	gptid/f6e3d186-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
  mirror-3									  ONLINE	   0	 0	 0
	gptid/4f8cc1aa-24b3-11e7-a0f7-000c296e056b  ONLINE	   0	 0	 0
	gptid/f81888ab-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
  mirror-4									  ONLINE	   0	 0	 0
	gptid/f8b3cf1f-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
	gptid/df5f5f29-a346-11e7-b49a-000c296e056b  ONLINE	   0	 0	 0
  mirror-5									  ONLINE	   0	 0	 0
	gptid/d6cc7c7e-533e-11e8-9b0b-000c296e056b  ONLINE	   0	 0	 0
	gptid/fa7efa09-6bed-11e6-985e-000c296e056b  ONLINE	   0	 0	 0
logs
  gptid/3a02af31-b913-11e6-823d-000c296e056b	FAULTED	  0	 8	 0  too many errors

 

[Chris Moore]

PS. The command you need should be zpool remove <ZPOOL> <DEVICE>, so for your pool, it would be:

zpool remove Jabba gptid/3a02af31-b913-11e6-823d-000c296e056b or something very similar. You can also do this from the GUI.
https://www.ixsystems.com/documentation/freenas/11/storage.html#removing-a-log-or-cache-device