Reset drive "critical" state

[ozon32]

Sorry, it's done in English now, thanks!

 

[Macmcd]

I apologise for reviving this thread but I am having a few issues trying to follow the instructions. I have a DS2419, running DSM 7.2.1-69057 Update 6, with a DX1215 attached. Both units are full of various sized IronWolf Pro drives and over the space of 24 hours, three of the drives have been marked as 'Critical'. All three drives are 'Normal' after an IronWolf Health check and 'Healthy' according to S.M.A.R.T. so, after reading the thread, I attempted the proposed solution. I have no knowledge of sqlite so have dutifully copied the commands substituting my info. When I get to:

DELETE FROM logs WHERE serial LIKE '[drive S/N]';

or

SELECT * FROM logs WHERE serial LIKE '[drive S/N]';

I receive the following error:

Parse error: no such table: logs

Could someone please point me in the right direction?

 

[posterberg]

I have more or less the same issue. I have two drives marked as crashed. One does indeed seem to be crashed, but the other one runs just fine when hooking it up directly to my desktop in an external disk cradle.
I have run a thorough S.M.A.R.T test and it is marked as healthy.

I really want to try to get this disk re-activated and have my RAID5 set rebuilt using the new disk that I have put in the slot where the really crashed disk was.

Problem is that I can't seem to be able to re-activate the healthy disk.

I have deleted any traces from it in .SYNODISKBD.

sqlite> delete from logs where serial='ZHZ547YT';
sqlite> select * from logs where serial = "ZHZ547YT";
sqlite>

Rebooted the nas but I still get it to show up as crashed. Is there something else that I need to do to force the disk to get re-activated?

 

[atakacs]

Did you have anything in the logs for this serial before the delete ?

Try to dump everything from the logs - something might point you in the right direction. Failing that try to delete the whole log (not great but worth a try).

 

[Telos]

Just throwin' this out there... may not be applicable.

 

[posterberg]

Did you have anything in the logs for this serial before the delete ?

Try to dump everything from the logs - something might point you in the right direction. Failing that try to delete the whole log (not great but worth a try).

There were several logs before I made the delete. I also made a backup of the sqlitedb-file before changing anything to the file so I can always go back if needed. I will try to delete the entire database and see if that helps.
Thanks!

 

[posterberg]

Just throwin' this out there... may not be applicable.

Seems like most of that is related to extX file systems. I am working with raid5 set so it wouldn't work to do a extfsck on a single disk. But I will try if I have any luck with synosetkeyvalue /etc/synoinfo.conf disable_volumes volume1

 

[posterberg]

There were several logs before I made the delete. I also made a backup of the sqlitedb-file before changing anything to the file so I can always go back if needed. I will try to delete the entire database and see if that helps.
Thanks!

sqlite> delete from logs;
sqlite> select * from logs;
sqlite>

Deleted everything from logs, made no difference.

Rebooted the NAS, still no difference. Are there really no backend command from SYNO that can force re-enabling a disk? What would third line support do?

 

[posterberg]

I also notice two more files along with the db file itself.

-rw-r--r--  1 system log    40960 Jun 19 18:59 .SYNODISKDB
-rw-r--r--  1 system log    32768 Jun 19 18:59 .SYNODISKDB-shm
-rw-r--r--  1 system log    37112 Jun 19 18:59 .SYNODISKDB-wal

I moved all three to a separate backup folder and restarted the NAS again. But still says that drive one is crashed, even though I can talk to it nicely from command line. No problems at all. So there has to exist some more place in the OS where the NAS keeps track of this.

I have noticed many files in /etc/ that I am curious to edit/remove /etc/space/space_table/space_table_<TIMESTAMP>.xml.
I am not sure why there are many of these files with similar content or what they are used for, but clearly my drive is marked as down.

[
    {
        "blk_luns": [],
        "can_assemble": false,
        "can_reboot_assemble": false,
        "dev_type": 2,
        "faulty_disks": [
            "ZHZ547YT"
        ],
        "id": "reuse_1",
        "limited_raidgroup_num": 12,
        "missing": true,
        "num_id": 1,
        "raids": [
            {
                "assemembled": false,
                "designed_disk_cnt": 12,
                "devs": [
                    {
                        "cnr_type": 1,
                        "container_id": 0,
                        "container_name": "",
                        "dev_path": "/dev/sda3",
                        "disk_order_id": 1,
                        "display_name": "Disk 1",
                        "drive_protocol_type": 0,
                        "drive_protocol_type_str": "SATA",
                        "drive_type": "hdd",
                        "failed": true,
                        "in_use": false,
                        "loc": "0-1",
                        "missing": false,
                        "model": "ST10000VN0008-2JJ101",
                        "pci_slot": -1,
                        "port_type": "normal",
                        "serial": "ZHZ547YT",
                        "slot": 0,
                        "total_size": 10000831348736,
                        "vendor": "Seagate"
                    },
...
...

I am quite curious to move all these files to some place where SYNO can't find them, and move them back if its unrelated. Any suggestions?

 

[Trollti]

I had to scrap from the critical list the SN of a 20TB Toshiba HDD and the sqlite method worked like a charm on my RS1219+.

Rebooted through the webUI, and my crtical state disappeared

The problem was an IO error... indeed I had physically extracted/inserted the same disk during a rebuild, which logically was a quite severe IO problem :/


Nice solution, quite obvious I imagine for those who studied SQL... always felt asleep studying that ^^

Trollti

PS: I always stay within the mdadm scope... soft raid rules (hence,... cannot say if works with SHR1/2)

 

[DSfuchs]

So there has to exist some more place in the OS where the NAS keeps track of this.

Yes, please see also:

Storage Pool Crash - Advice Needed

Hello all, My RS1221+ suffered a Storage Pool crash today so look for advice on how to recover it as the symptoms of this one are new to me. Configuration: Storage Pool 1, Volume 1 = 6 x SSDs in SHR1 Storage Pool 2, Volume 2 = 2 x HDDs in SHR (mirrored backup of Volume 1) Event: Overnight...

www.synoforum.com

 

[DSfuchs]

or the new Resource here:

Prepared for HDD/SSD issues.

Sooner or later, sometimes even sooner than that, problems involving HDDs or SSDs are bound to arise within the Synology Storage Manager. However, critical errors are not always attributable to a hardware-related cause. To help you narrow down...

www.synoforum.com

 

[Andromeda]

does this no longer apply to the most recent dsm version?

See this post. Quite nice!