ZFS Changed HBA, ZFS now carps about block size

scotia

scotia

Hi all,

I recently changed my HBA/controller in my NAS from an Areca to an LSI.

Upon doing so, zpool status now says the block sizes mismatch.

Here's the pool on the Areca:
Code: 
NAME                                            STATE     READ WRITE CKSUM
ZP-02                                           ONLINE       0     0     0
  raidz1-0                                      ONLINE       0     0     0
    gptid/62c9ccd0-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0
    gptid/634a1d39-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0
    gptid/6366d6e0-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0
    gptid/638a9979-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0
    gptid/63a88743-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0
    gptid/6425af40-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0

And on the LSI:
Code: 
NAME                                            STATE     READ WRITE CKSUM
ZP-02                                           ONLINE       0     0     0
  raidz1-0                                      ONLINE       0     0     0
    gptid/62c9ccd0-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0  block size: 512B configured, 4096B native 
    gptid/634a1d39-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0  block size: 512B configured, 4096B native 
    gptid/6366d6e0-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0  block size: 512B configured, 4096B native 
    gptid/638a9979-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0  block size: 512B configured, 4096B native 
    gptid/63a88743-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0  block size: 512B configured, 4096B native 
    gptid/6425af40-f906-11ea-8219-3cd92b042b6f  ONLINE       0     0     0  block size: 512B configured, 4096B native 
cache
  gpt/zp-02-cache                               ONLINE       0     0     0
(Yes, I added a cache).

Is this expected?

Thanks.
 
Interesting that it started showing up after changing controller. I assume the LSI must be more transparent regarding disk geometry/etc.

It will require re-creating the pool to change and likely isn't a huge issue unless you're after every mb of performance so I would probably just ignore it.
 
usdmatt said:
It will require re-creating the pool
Click to expand...

That's not going to happen :). It's holding 30TB and I don't have a spare 30TB to put it...

It's all large media files, written once, so the performance hit shouldn't be noticeable if my understanding is correct.
 
You must log in or register to reply here.
Back
Top