Checksum Errors while Resilvering

TrueNAS General
, ,
1

I am upgrading my pool by adding in 26tb drivesw to replace my 16 tb ones. This is a Raidz2 pool with 6 drives, each drive takes about 24 hours to resilver. I just received two checksum errors on the resilvering process, zero errors on the disk that I am adding in. Is this a concern, or will the resilver correct the error? Thanks

2

What does zpool status -v say?

3

26 TB drives? No typo? And, most critically, no SMR?

4

Exactly what is the Model of the drives you are using?

Also, have you verified that these are not the Chinese drives which flooded the market recently as new drives? Read the label, read the SMART data, verify the warranty as well.

5

It shows :

    NAME                                        STATE     READ WRITE CKSUM
    Centerville                                 DEGRADED     0     0     0
      raidz2-0                                  DEGRADED     0     0     0
        72ad2ea2-3789-48c6-9f1a-92fd5e8e4aa0    ONLINE       0     0     0
        replacing-1                             DEGRADED     0     0     2
          4795428958891218551                   UNAVAIL      0     0     0  was /dev/disk/by-partuuid/54e88a98-a2b6-4295-8cfa-c82c2daf5768
          6a13d287-1f5f-4daa-8f71-b4c0cc83e4f3  ONLINE       0     0     0  (resilvering)
        9eea628f-6099-4e85-a55c-4179cef75d87    ONLINE       0     0     3  (resilvering)
        c726f6b3-35be-4aa3-8909-81ed1e4b0d7a    ONLINE       0     0     0
        bd9a979a-3994-4835-b71a-e83ea64c4b24    ONLINE       0     0     0
        b096a22d-ac13-46a6-b26c-ed54401d641b    ONLINE       0     0     0

I just noticed that a second drive also has developed errors and is alos resilvering. Perhaps these new drives have defects?

6

These are Seagate EXOS ST28000NM000C Drives purchsed from Server Parts Deals in the usa and are supposed to be manufacturer refurbished and carry a 2 year warranty. This is my first experience with them. And I am becoming concerned that two of them have reported checksum errors.

7

I can’t say that I have heard anything bad about them and they have been around for a while. The only thing I thought odd was a New drive is $799.99 (3 year warranty), Manufacturer Refurb is $364.99 (2 year warranty), Seller Refurb is $779.99 (90 day warranty). Those sound wrong. Of course I would take the Manufacturer Refurb units all day long at those prices.

I’m sure if someone had a bad experience, they will post it.

Did you run BADBLOCKS or some sort of test on these drives before putting them into service? It is always prudent to do so. With these large capacities these tests take forever!

Are you replacing one drive at a time, waiting for it to complete the Resilver before installing a second drive? Explain what you are doing, be specific.

1 Like
8

HI… I am resilvering them one at a time. Haven’t run any disnostics, I want to let this resilver finish before I check out the cause. But you are right I should have run some diagnostics on them first. I did not.

9

For interest, when iot was done resilvering my drive, I rebooted my server and all the checksums went away. I am not sure if this means the problem was resolved, or if the system just cleared the error without it being fixed. I will finish my resilvering to complete the capacity upgrade then run a scrub to see if anything shows up… Tanks everybody for giving me adivce. It is one of the things I like the best about the Truenas community.