RAID5 degraded - Sehr seltsames Verhalten

Hallo,
gestern Abend ist mir etwas extrem seltsames passiert: Ich war am Dartspielen, als plötzlich mein TS-469 Pro Alarmtöne von sich gab. Ich habe dann gleich auf die Weboberfläche geschaut, wo ich folgende Eintrage im Log finden konnte:

Unbenannt.PNG

Ein Neustart hat nichts gebracht und beim Versuch des Widerherstellens, bekomme ich nur die Fehlermeldung, dass das nicht möglich sei. Anscheinend gab es dieses Problem bei anderen auch schon: http://forum.qnap.com/viewtopic.php?t=110067

Das ist die Ausgabe von demsg:

40.073341] md: unbind<sdc3>
[   40.081284] md: export_rdev(sdc3)
[   40.082355] md/raid:md1: device sda3 operational as raid disk 0
[   40.082361] md/raid:md1: device sdb3 operational as raid disk 1
[   40.093369] md/raid:md1: allocated 68992kB
[   40.093471] md/raid:md1: not enough operational devices (2/4 failed)
[   40.093497] RAID conf printout:
[   40.093500]  --- level:5 rd:4 wd:2
[   40.093506]  disk 0, o:1, dev:sda3
[   40.093510]  disk 1, o:1, dev:sdb3
[   40.100678] md/raid:md1: failed to run raid set.
[   40.100684] md: pers->run() failed ...
[   40.100746] md: md1 stopped.
[   40.100760] md: unbind<sda3>
[   40.106292] md: export_rdev(sda3)
[   40.106366] md: unbind<sdb3>
[   40.114285] md: export_rdev(sdb3)
[   40.222498]  md13: unknown partition table
[   65.193026] kjournald starting.  Commit interval 5 seconds
[   65.463843] EXT3-fs (md13): using internal journal
[   65.463850] EXT3-fs (md13): mounted filesystem with ordered data mode
[   84.390566] cryptodev: driver 0.9 loaded.
[   84.413318] Register ICH/PCH RTC driver.
[   84.415683] iTCO_vendor_support: vendor-support=0
[   84.417823] iTCO_wdt: Intel TCO Watch

Wer kann mir helfen? Vielen Dank

Von den wichtigsten Daten natürlich Trotzdem etwas ärgerlich.

Vorallem scheint das schon öfter mal vorgekommen sein, wenn ich das mal Google.

Keines Falls

Meine Platten sind ja vollkommen in Ordnung. - Es wurde kurzzeitig erkannt, dass ich 2 Platten entfernt hätte, was ich nicht habe. Die HDDs selbst zeigen keinerlei Probleme.

Hab dem Support schon mal geschrieben, mal sehen was die sagen. Interssant finde ich, dass es mehrere Forenbeiträge in anderen Foren gibt, die exakt das gleiche Problem haben. Dort konnte dann der Support helfen.

Bin ja mal echt gespannt was da los ist...

WD RED 4TB NASware 3.0

Der hat genau das selbe Problem. selbe statusausgaben etc.: http://forum.qnap.com/viewtopic.php?f=25&t=110067&start=15

dort konnte der support das Problem fixen.

Support ist doch Support?

antwort kam am selben tag abends.
wollte details zu hdds und logs haben

Also bislang stellt der Support auf dumm...
Bin inzwischen total genervt.

Jemand einen Ansprechpartner?

Hier das Statement vom UK Support. Jetzt natürlich meine Frage, wie kann sowas passieren? Und warum wird meine USV als nicht kompatibel gelistet?

Zitat

Dear Florian

Thanks for writing us.

Sorry for the inconvenience caused.

We have checked your kernel logs thoroughly, we found the drives been failed unfortunately.

As you know that Raid-5 can sustain with 1 drive failure but in your case 3 drives (Drive 2,3,4) were having issues in Raid-5. Out of 3 drives, 2 drives (Drive 3 & 4) failed completely.

================================================================
Initially Drive 4, 3 and 2 are having issues.

Logs snippet:
4408, 0,2016-06-06,22:48:26,System,127.0.0.1,localhost,Host: Drive4 Disabled NCQ since timeout error., 0,1465246104,
4409, 0,2016-06-06,22:48:28,System,127.0.0.1,localhost,Host: Drive3 Disabled NCQ since timeout error., 0,1465246108,
4410, 1,2016-06-06,22:48:30,System,127.0.0.1,localhost,Host: Drive2 Read I/O error, sense_key=0x0, asc=0x0, ascq=0x0, CDB=88 00 00 00 00 01 70 67 89 58 00 00 01 30 00 00 ., 0,1465246110,

Then Drive 3 failed because of this reason, Raid-5 became Degraded.

Logs snippet:

4414, 2,2016-06-06,22:48:55,System,127.0.0.1,localhost,Hot-remove Host: Drive3 failed., 0,1465246135,
4415, 1,2016-06-06,22:48:56,System,127.0.0.1,localhost,[Pool 1] RAID Group 1 is in degraded mode., 0,1465246136,

After then, Drive 4 with lots of Write I/O hence Raid-5 group cannot read and write on this drive as well hence system automatically detected this condition and mounted this Raid group as "degraded and read-only mode." . This is Special condition, wherein you cannot recover the Raid but system let the user to take only the backup.

Logs snippet:

4422, 1,2016-06-06,22:49:13,System,127.0.0.1,localhost,[Pool 1] RAID Group 1 is in degraded and read-only mode., 0,1465246153,
4423, 1,2016-06-06,22:49:15,System,127.0.0.1,localhost,[Volume RAID5, Pool 1] Mount the file system read-only., 0,1465246155,

Unfortunately drive 4 also completely failed so Raid-5 cannot handle 2 drive failure hence it became inactive mode.

Logs snippet:
4430, 2,2016-06-06,23:03:31,System,127.0.0.1,localhost,[Pool 1] RAID Group 1 is inactive., 0,1465247011,
=========================================================================

In this case, 3 drives are having issues. Raid-5 can sustain one drive failure but 2 drive failed and 1 drive error out. Sorry, it is not possible to recover and you have to replace those drives, delete the config and start from scratch.

HDD smart test cannot test the SMART values of the drives, if the drives are not completely accessible because of I/O error. You can do the scan drive on those disks, it can show the real identity of the drive status.
If not, you can test with HDD vendor tools to know the exact status of the drives.

WD Data Lifeguard Diagnostics : http://support.wdc.com/KnowledgeBase/answer.aspx?ID=940

Seagate Sea tools: http://knowledge.seagate.com/articles/en_US/FAQ/202435en

Note: one more suggestion, please use the compatible UPS to your NAS model. Your exisiting UPS "American Power Conversion Back-UPS BR 900GI" device is not compatible to your NAS model.
Please find the compatibility list here: https://www.qnap.com/i/in/compatibility/?model=21&category=9

If you have queries, please feel free to write us. Thanks!

Alles anzeigen

Zitat

Dear Florian

Sometimes the 2 or more drives can fail at same time. Its a rare and unfortunate situation. This is not a backplane issue from our inspection on the logs.

This is more related to drive reliability issues especially WD RED drives and seagate st3000dm001/st2000dm001 drives. We have encountered lots of drive reliability issues in our experience.

You can use alternatively, other drive models from WD vendor like WD4000F9YZ, WD4000FYYZ etc.

You can check with HDD vendor tools to know about the results of these drives.

UPS incompatibility issue is nothing to do with your drive failures. As we checked your logs, your UPS was logged in and out randomly. When we checked your UPS model, it was not in your list hence we suggested you to use the compatible one.

Alles anzeigen