Results 1 to 2 of 2

Thread: Slow response - RAID problem?

  1. #1
    Join Date
    Feb 2008
    Beans
    119

    Slow response - RAID problem?

    I'm getting slow/delayed response on a Kubuntu 19.04 system. The home directory is on a software two disk RAID1, see below. I see a lot of errors on one of the RAID disks, and resync is taking forever, also see below. My questions are:

    -Is it likely that the sdb errors are what's causing delayed response, and/or how do I further diagnose?

    -If the disk is the problem, I'm perfectly willing to replace the disk, or both disks for that matter, since they're both the same age, but ... are there things
    I should do/try first?

    I'm somewhere between a novice and an expert, but this problem is at the edge of my capabilities to diagnose/fix, so recommendations to diagnose and fix appreciated .... Gus

    $ mdadm --detail /dev/md0
    /dev/md0:
    Version : 1.2
    Creation Time : Fri Aug 19 08:34:54 2016
    Raid Level : raid1
    Array Size : 1829867904 (1745.10 GiB 1873.78 GB)
    Used Dev Size : 1829867904 (1745.10 GiB 1873.78 GB)
    Raid Devices : 2
    Total Devices : 2
    Persistence : Superblock is persistent

    Intent Bitmap : Internal

    Update Time : Fri Sep 13 11:21:49 2019
    State : active, resyncing
    Active Devices : 2
    Working Devices : 2
    Failed Devices : 0
    Spare Devices : 0

    Consistency Policy : bitmap

    Resync Status : 0% complete

    Name : Jboat17:0 (local to host Jboat17)
    UUID : cfde6372:28fef717:4d35e851:bbf99b6b
    Events : 87151

    Number Major Minor RaidDevice State
    0 8 17 0 active sync /dev/sdb1
    1 8 1 1 active sync /dev/sda1


    And I'm getting a lot of these ............

    $ tail -f /var/log/kern.log

    Sep 13 11:12:44 Jboat17 kernel: [ 2585.835973] ata2.00: exception Emask 0x0 SAct 0x407400 SErr 0x0 action 0x0
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.835980] ata2.00: irq_stat 0x40000008
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.835986] ata2.00: failed command: READ FPDMA QUEUED
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.835997] ata2.00: cmd 60/00:60:80:75:79/05:00:00:00:00/40 tag 12 ncq dma 655360 in
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.835997] res 41/40:00:70:78:79/00:00:00:00:00/40 Emask 0x409 (media error) <F>
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.836001] ata2.00: status: { DRDY ERR }
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.836005] ata2.00: error: { UNC }
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837278] ata2.00: configured for UDMA/133
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837330] sd 1:0:0:0: [sdb] tag#12 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837335] sd 1:0:0:0: [sdb] tag#12 Sense Key : Medium Error [current]
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837339] sd 1:0:0:0: [sdb] tag#12 Add. Sense: Unrecovered read error - auto reallocate failed
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837344] sd 1:0:0:0: [sdb] tag#12 CDB: Read(10) 28 00 00 79 75 80 00 05 00 00
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837348] print_req_error: I/O error, dev sdb, sector 7960688 flags 4000
    Sep 13 11:12:44 Jboat17 kernel: [ 2585.837376] ata2: EH complete

    And resync is taking forever ....

    $ sudo watch cat /proc/mdstat

    Every 2.0s: cat /proc/mdstat
    Personalities : [raid1] [linear] [multipath] [raid0] [raid6] [raid5] [raid4] [raid10]
    md0 : active raid1 sda1[1] sdb1[0]
    1829867904 blocks super 1.2 [2/2] [UU]
    [>....................] resync = 0.3% (5731456/1829867904) finish=7975.9min speed=3811K/sec
    bitmap: 14/14 pages [56KB], 65536KB chunk

    unused devices: <none>

  2. #2
    Join Date
    Mar 2010
    Location
    Squidbilly-Land
    Beans
    Hidden!
    Distro
    Ubuntu

    Re: Slow response - RAID problem?

    Looks like a bad cable on one of the disks or a failing controller port. I'd take the array off line. It looks likely that corrupted data is being written.

    Raw SMART data would be helpful on both disks. smartctl is the tool.
    Step 1 - run the tests on each disk first. Wait for the tests to finish.
    Step 2 - run the report on each disk. Look at the raw data.

    You can lookup each of the SMART values for what they mean.


    Oh and if you don't already have backups, YOU NEED SOME ASAP.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •