Upgrade 9.1.0 block alignment Warning

[hellweiss]

Hi,

after upgrading, I see this new message: (zpool is not upgraded yet.)

WARNING: Disk, '/dev/gptid/24ae079b-ae5a-11e1-8093-90e2ba0d08e8', has a block alignment that is larger than the pool's alignment
WARNING: Disk, '/dev/gptid/2635dd13-ae5a-11e1-8093-90e2ba0d08e8', has a block alignment that is larger than the pool's alignment

These are the LOG Disks:

logs
mirror-2 ONLINE
gptid/24ae079b-ae5a-11e1-8093-90e2ba0d08e8 ONLINE
gptid/2635dd13-ae5a-11e1-8093-90e2ba0d08e8 ONLINE

Something I can do about it ?

 

[William Grzybowski]

Possibly because your disks have a sector size of 4k and your pool is 512, which would mean bad performance.

Can you paste "zdb | grep ashift" and "diskinfo -v ada1"

 

[hellweiss]

zdb|grep ashift
ashift: 9
ashift: 9
ashift: 9

da1
512 # sectorsize
599988895744 # mediasize in bytes (558G)
1171853312 # mediasize in sectors
0 # stripesize
0 # stripeoffset
72944 # Cylinders according to firmware.
255 # Heads according to firmware.
63 # Sectors according to firmware.
LXVTP82MCC83C7003D1B # Disk ident.

Thank's

 

[William Grzybowski]

Well, thats weird, everything looks in place. I cannot tell you why of this warning message.

 

[hellweiss]

Perhaps it's encryption needs 4K Blocksize ?!?

GEOM_ELI: Device da23p1.eli created.
GEOM_ELI: Encryption: AES-XTS 256
GEOM_ELI: Crypto: software
Its directly after this its da24 and da25

 

[William Grzybowski]

I dont think so, the message seems related to ZFS pool, and the encryption is used for the swap.

 

[hellweiss]

Performance is great so I'll leave it for now.

Thanks for your help.

 

[cyberjock]

hellweiss, what's the model and brand of the hard drive that da1 is? I bet your disk is one of those 4k sector disks that lies and says its 512 bytes.

 

[pintapiva]

Hi all! I have this message in my logs too.

Code:

Solaris: WARNING: Disk, '/dev/gptid/a66c4ea7-99fd-11e2-9a67-00241dc2d7bf', has a block alignment that is larger than the pool's alignment
 
Solaris: WARNING: Disk, '/dev/gptid/a66c4ea7-99fd-11e2-9a67-00241dc2d7bf', has a block alignment that is larger than the pool's alignment
 
Solaris: WARNING: Disk, '/dev/gptid/a66c4ea7-99fd-11e2-9a67-00241dc2d7bf', has a block alignment that is larger than the pool's alignment

Code:

#zdb | grep ashift
ashift : 12

Code:

[root@freenas] /# diskinfo -v ada1
ada1
    512            # sectorsize
    2000398934016    # mediasize in bytes (1.8T)
    3907029168      # mediasize in sectors
    4096            # stripesize
    0              # stripeoffset
    3876021        # Cylinders according to firmware.
    16              # Heads according to firmware.
    63              # Sectors according to firmware.
    WD-WMC300546910    # Disk ident.
 
[root@freenas] /# diskinfo -v ada2
ada2
    512            # sectorsize
    2000398934016    # mediasize in bytes (1.8T)
    3907029168      # mediasize in sectors
    4096            # stripesize
    0              # stripeoffset
    3876021        # Cylinders according to firmware.
    16              # Heads according to firmware.
    63              # Sectors according to firmware.
    WD-WMC300548205    # Disk ident.
 
[root@freenas] /# diskinfo -v ada3
ada3
    512            # sectorsize
    2000398934016    # mediasize in bytes (1.8T)
    3907029168      # mediasize in sectors
    4096            # stripesize
    0              # stripeoffset
    3876021        # Cylinders according to firmware.
    16              # Heads according to firmware.
    63              # Sectors according to firmware.
    WD-WMC300558688    # Disk ident.
 
[root@freenas] /# diskinfo -v ada4
ada4
    512            # sectorsize
    2000398934016    # mediasize in bytes (1.8T)
    3907029168      # mediasize in sectors
    4096            # stripesize
    0              # stripeoffset
    3876021        # Cylinders according to firmware.
    16              # Heads according to firmware.
    63              # Sectors according to firmware.
    WD-WMC300548290    # Disk ident.

and in log I have a one mo strange line:

Code:

KLD geom_raid5.ko: depends on kernel - not available or version mismatch
linker_load_file: Unsupported file type

but all pools works well

 

[pbucher]

Ditto I'm seeing it on my 9.1 boxes too, I'm pretty sure the message wasn't there with 9.1RC2 though I could have missed it. Definitely wasn't there for RC1.

My drives are Seagate ST1000NM0001's and I pulled the spec sheet for them which listed 512,520,528 for bytes per sector.

[root@sandbox] ~# diskinfo -v /dev/da2
/dev/da2
512 # sectorsize
1000204886016 # mediasize in bytes (931G)
1953525168 # mediasize in sectors
0 # stripesize
0 # stripeoffset
121601 # Cylinders according to firmware.
255 # Heads according to firmware.
63 # Sectors according to firmware.
Z1N3VKZH0000C3337NTH # Disk ident.

 

[Quadgnim]

I'm seeing it now, and all 4 of my drives are the same brand,

Seagate, SV35 Series ST2000VX000 2 TB 7200 RPM drives. I have 4 brand new drives. Do I need to be worried about this? Is this something I can fix, by recreating my pools with a special command? I just created these brand new pools, so I'm concerned, I don't want to have to move the data, again...

 

[Quadgnim]

I have more to add. I'm not sure if it's related or not, but it's killing me here. Any help is greatly appreciated. The same drive that keeps reporting it has the wrong alignment, also keeps being evicted:

Aug 23 10:00:24 freenas kernel:
Aug 23 10:02:20 freenas manage.py: [common.pipesubr:57] Popen()ing: /usr/local/bin/warden list -v
Aug 23 10:02:21 freenas last message repeated 2 times
Aug 23 10:18:44 freenas afpd[3693]: AFP3.3 Login by timemachine
Aug 23 10:18:45 freenas afpd[3693]: AFP logout by timemachine
Aug 23 10:18:45 freenas afpd[3693]: AFP statistics: 0.85 KB read, 0.81 KB written
Aug 23 10:18:45 freenas afpd[3693]: done
Aug 23 10:19:07 freenas afpd[3705]: AFP3.3 Login by timemachine
Aug 23 10:29:45 freenas kernel: (ada3:ahcich3:0:0:0): lost device
Aug 23 10:29:46 freenas kernel: (ada3:ahcich3:0:0:0): removing device entry
Aug 23 10:29:49 freenas kernel: ada3 at ahcich3 bus 0 scbus3 target 0 lun 0
Aug 23 10:29:49 freenas kernel: ada3: ATA-8 SATA 3.x device
Aug 23 10:29:49 freenas kernel: ada3: 300.000MB/s transfers (SATA 2.x, UDMA6, PIO 8192bytes)
Aug 23 10:29:49 freenas kernel: ada3: Command Queueing enabled
Aug 23 10:29:49 freenas kernel: ada3: 1907729MB (3907029168 512 byte sectors: 16H 63S/T 16383C)
Aug 23 10:29:49 freenas kernel: ada3: Previously was known as ad10
Aug 23 10:32:39 freenas afpd[3705]: AFP logout by timemachine
Aug 23 10:32:39 freenas afpd[3705]: AFP statistics: 3093893.05 KB read, 304093.11 KB written
Aug 23 10:32:39 freenas afpd[3705]: done

I thought I had a bad drive as it kept happening. If I reboot it comes back but as soon as I put a load on the pool I lose it again. After replacing it with a brand new drive I'm still getting the same error.

Then I thought maybe it was a bad cable so I replaced the cable, still no improvement.

Is this a HW issue, if so it would have to be the mother board, but I've been using this MB for around a year and this issue only just started.

Any other ideas?

Thanks

Larry

 

[pbucher]

I'm seeing it now, and all 4 of my drives are the same brand,

Seagate, SV35 Series ST2000VX000 2 TB 7200 RPM drives. I have 4 brand new drives. Do I need to be worried about this? Is this something I can fix, by recreating my pools with a special command? I just created these brand new pools, so I'm concerned, I don't want to have to move the data, again...

The alignment errors appear to be incorrect for most people and are harmless for the actual operation of system. Even if you have 4k drives and they are unaligned you still shouldn't suffer anything more then bad performance.

 

[pintapiva]

I have more to add. I'm not sure if it's related or not, but it's killing me here. Any help is greatly appreciated. The same drive that keeps reporting it has the wrong alignment, also keeps being evicted:

Aug 23 10:00:24 freenas kernel:
Aug 23 10:02:20 freenas manage.py: [common.pipesubr:57] Popen()ing: /usr/local/bin/warden list -v
Aug 23 10:02:21 freenas last message repeated 2 times
Aug 23 10:18:44 freenas afpd[3693]: AFP3.3 Login by timemachine
Aug 23 10:18:45 freenas afpd[3693]: AFP logout by timemachine
Aug 23 10:18:45 freenas afpd[3693]: AFP statistics: 0.85 KB read, 0.81 KB written
Aug 23 10:18:45 freenas afpd[3693]: done
Aug 23 10:19:07 freenas afpd[3705]: AFP3.3 Login by timemachine
Aug 23 10:29:45 freenas kernel: (ada3:ahcich3:0:0:0): lost device
Aug 23 10:29:46 freenas kernel: (ada3:ahcich3:0:0:0): removing device entry
Aug 23 10:29:49 freenas kernel: ada3 at ahcich3 bus 0 scbus3 target 0 lun 0
Aug 23 10:29:49 freenas kernel: ada3: ATA-8 SATA 3.x device
Aug 23 10:29:49 freenas kernel: ada3: 300.000MB/s transfers (SATA 2.x, UDMA6, PIO 8192bytes)
Aug 23 10:29:49 freenas kernel: ada3: Command Queueing enabled
Aug 23 10:29:49 freenas kernel: ada3: 1907729MB (3907029168 512 byte sectors: 16H 63S/T 16383C)
Aug 23 10:29:49 freenas kernel: ada3: Previously was known as ad10
Aug 23 10:32:39 freenas afpd[3705]: AFP logout by timemachine
Aug 23 10:32:39 freenas afpd[3705]: AFP statistics: 3093893.05 KB read, 304093.11 KB written
Aug 23 10:32:39 freenas afpd[3705]: done

I thought I had a bad drive as it kept happening. If I reboot it comes back but as soon as I put a load on the pool I lose it again. After replacing it with a brand new drive I'm still getting the same error.

Then I thought maybe it was a bad cable so I replaced the cable, still no improvement.

Is this a HW issue, if so it would have to be the mother board, but I've been using this MB for around a year and this issue only just started.

Any other ideas?

Thanks

Larry

Hello! I think it is necessary to replace a hard drive.

 

[Visseroth]

Well add me to the list of people having this problem too. I just started getting email notifications of my pool giving me the same errors only it's ALL of my drives......

Code:

+++ /tmp/security.1gYA85NV    2013-12-08 03:01:01.000000000 -0800
+Solaris: WARNING: Disk, '/dev/gptid/a93ca521-2e05-11e2-94a7-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/0901b3e1-c196-11e2-9bcc-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/886b66c0-3168-11e2-91cf-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/5173c89c-2ece-11e2-b465-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/426780f3-305c-11e2-82a9-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/7037831f-1829-11e2-8c8f-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/719e06f4-2e06-11e2-94a7-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/47688090-2f6d-11e2-bef9-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/d1fb51e4-325d-11e2-9ff7-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/10373ccc-2ecf-11e2-b465-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/d04581ae-305c-11e2-82a9-003048348d66', has a block alignment that is larger than the pool's alignment
+
+Solaris: WARNING: Disk, '/dev/gptid/5db2e52b-316b-11e2-91cf-003048348d66', has a block alignment that is larger than the pool's alignment
+

Code:

zdb | grep ashift
            ashift: 9
            ashift: 9

Code:

diskinfo -v ada1
ada1
        512            # sectorsize
        2000398934016  # mediasize in bytes (1.8T)
        3907029168      # mediasize in sectors
        4096            # stripesize
        0              # stripeoffset
        3876021        # Cylinders according to firmware.
        16              # Heads according to firmware.
        63              # Sectors according to firmware.
        S2H7JX0CA00492  # Disk ident.

 

[Stephen J]

Me too. All 10 disks in my array.