WEHA Posted June 28, 2017 Share Posted June 28, 2017 (edited) So I installed a new unraid. Assigned all disks for the array and cache. Everything was working.. until I rebooted the server. Array was fine but cache disk showed with a blue icon (new disk). I've tried (i think) everything (after manually copying the data back to the array): - format disks - wipefs - preclear - new config It all stays the same, disks with blue icons and no way to make it a cache pool. When I click format it says "formatting" once but when reloading the main page it's possible to click format again. Any help would be appreciated I tried a lot of searches but came up empty. thank you Edited July 8, 2017 by WEHA Quote Link to comment
hernandito Posted June 28, 2017 Share Posted June 28, 2017 (edited) I think once you assign the parity drive, check ok to format drive and start the array and it will do it automatically (the formatting) don't use unassigned devices plugin for any of this. Edited June 28, 2017 by hernandito Quote Link to comment
WEHA Posted June 29, 2017 Author Share Posted June 29, 2017 I'm not sure what you mean, the array is fine. It's the cache pool that is not working. I did use "was" fine but it should read "is" fine I've attached a screenshot to illustrate the problem Quote Link to comment
WEHA Posted June 29, 2017 Author Share Posted June 29, 2017 attached diagnostics-20170629-1527.zip Quote Link to comment
JorgeB Posted June 29, 2017 Share Posted June 29, 2017 Stopt the array and try this (in this order) on the console/ssh: (if you rebooted check to see if the cache devices are still the same) wipefs -a /dev/sdg1 wipefs -a /dev/sdg wipefs -a /dev/sdf1 wipefs -a /dev/sdf Then start the array and try to format again. Quote Link to comment
WEHA Posted June 29, 2017 Author Share Posted June 29, 2017 wipefs -a /dev/sdg1 /dev/sdg1: 8 bytes were erased at offset 0x00010040 (btrfs): 5f 42 48 52 66 53 5f 4d wipefs -a /dev/sdg /dev/sdg: 2 bytes were erased at offset 0x000001fe (dos): 55 aa /dev/sdg: calling ioctl to re-read partition table: Success wipefs -a /dev/sdf1 wipefs -a /dev/sdf /dev/sdf: 2 bytes were erased at offset 0x000001fe (dos): 55 aa /dev/sdf: calling ioctl to re-read partition table: Success Executed, same result Log: Jun 29 17:24:24 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 17:24:24 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 17:24:24 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 17:24:24 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 17:24:24 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 17:24:24 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 17:24:24 unNeptunus emhttp: shcmd (2222): set -o pipefail ; mkfs.btrfs -f -K -dsingle -msingle /dev/sdg1 |& logger Jun 29 17:24:25 unNeptunus root: btrfs-progs v4.7.2 Jun 29 17:24:25 unNeptunus root: See http://btrfs.wiki.kernel.org for more information. Jun 29 17:24:25 unNeptunus root: Jun 29 17:24:25 unNeptunus root: Label: (null) Jun 29 17:24:25 unNeptunus root: UUID: 093ff426-5de6-46b3-ac66-46b4f70b8314 Jun 29 17:24:25 unNeptunus root: Node size: 16384 Jun 29 17:24:25 unNeptunus root: Sector size: 4096 Jun 29 17:24:25 unNeptunus root: Filesystem size: 1.82TiB Jun 29 17:24:25 unNeptunus root: Block group profiles: Jun 29 17:24:25 unNeptunus root: Data: single 8.00MiB Jun 29 17:24:25 unNeptunus root: Metadata: single 8.00MiB Jun 29 17:24:25 unNeptunus root: System: single 4.00MiB Jun 29 17:24:25 unNeptunus root: SSD detected: no Jun 29 17:24:25 unNeptunus root: Incompat features: extref, skinny-metadata Jun 29 17:24:25 unNeptunus root: Number of devices: 1 Jun 29 17:24:25 unNeptunus root: Devices: Jun 29 17:24:25 unNeptunus root: ID SIZE PATH Jun 29 17:24:25 unNeptunus root: 1 1.82TiB /dev/sdg1 Jun 29 17:24:25 unNeptunus root: Jun 29 17:24:25 unNeptunus emhttp: check_pool: /sbin/btrfs filesystem show /dev/sdg1 2>&1 Jun 29 17:24:25 unNeptunus kernel: BTRFS: device fsid 093ff426-5de6-46b3-ac66-46b4f70b8314 devid 1 transid 5 /dev/sdg1 Jun 29 17:24:25 unNeptunus emhttp: cacheUUID: Jun 29 17:24:25 unNeptunus emhttp: cacheNumDevices: 2 Jun 29 17:24:25 unNeptunus emhttp: cacheTotDevices: 0 Jun 29 17:24:25 unNeptunus emhttp: cacheNumMissing: 0 Jun 29 17:24:25 unNeptunus emhttp: cacheNumMisplaced: 0 Jun 29 17:24:25 unNeptunus emhttp: cacheNumExtra: 2 Jun 29 17:24:25 unNeptunus emhttp: shcmd (2232): mkdir -p /mnt/cache Jun 29 17:24:25 unNeptunus emhttp: mount error: No file system (no btrfs UUID) Jun 29 17:24:25 unNeptunus emhttp: shcmd (2233): umount /mnt/cache |& logger Jun 29 17:24:25 unNeptunus root: umount: /mnt/cache: not mounted Jun 29 17:24:25 unNeptunus emhttp: shcmd (2234): rmdir /mnt/cache Quote Link to comment
JorgeB Posted June 29, 2017 Share Posted June 29, 2017 Strange, can you repeat the wipe on both and try to add just one of the devices, start with sdg Quote Link to comment
WEHA Posted June 29, 2017 Author Share Posted June 29, 2017 Same result sdg: Jun 29 18:01:03 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 18:01:03 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 18:01:03 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 18:01:03 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 18:01:03 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 18:01:03 unNeptunus rc.diskinfo[8978]: SIGHUP ignored - already refreshing disk info. Jun 29 18:01:03 unNeptunus emhttp: shcmd (2453): set -o pipefail ; mkfs.btrfs -f -K -dsingle -msingle /dev/sdg1 |& logger Jun 29 18:01:04 unNeptunus root: btrfs-progs v4.7.2 Jun 29 18:01:04 unNeptunus root: See http://btrfs.wiki.kernel.org for more information. Jun 29 18:01:04 unNeptunus root: Jun 29 18:01:04 unNeptunus root: Label: (null) Jun 29 18:01:04 unNeptunus root: UUID: f9d1f8b9-2e2f-44a8-b3af-7148521794b3 Jun 29 18:01:04 unNeptunus root: Node size: 16384 Jun 29 18:01:04 unNeptunus root: Sector size: 4096 Jun 29 18:01:04 unNeptunus root: Filesystem size: 1.82TiB Jun 29 18:01:04 unNeptunus root: Block group profiles: Jun 29 18:01:04 unNeptunus root: Data: single 8.00MiB Jun 29 18:01:04 unNeptunus root: Metadata: single 8.00MiB Jun 29 18:01:04 unNeptunus root: System: single 4.00MiB Jun 29 18:01:04 unNeptunus root: SSD detected: no Jun 29 18:01:04 unNeptunus root: Incompat features: extref, skinny-metadata Jun 29 18:01:04 unNeptunus root: Number of devices: 1 Jun 29 18:01:04 unNeptunus root: Devices: Jun 29 18:01:04 unNeptunus root: ID SIZE PATH Jun 29 18:01:04 unNeptunus root: 1 1.82TiB /dev/sdg1 Jun 29 18:01:04 unNeptunus root: Jun 29 18:01:04 unNeptunus emhttp: check_pool: /sbin/btrfs filesystem show /dev/sdg1 2>&1 Jun 29 18:01:04 unNeptunus kernel: BTRFS: device fsid f9d1f8b9-2e2f-44a8-b3af-7148521794b3 devid 1 transid 5 /dev/sdg1 Jun 29 18:01:04 unNeptunus emhttp: cacheUUID: Jun 29 18:01:04 unNeptunus emhttp: cacheNumDevices: 1 Jun 29 18:01:04 unNeptunus emhttp: cacheTotDevices: 0 Jun 29 18:01:04 unNeptunus emhttp: cacheNumMissing: 0 Jun 29 18:01:04 unNeptunus emhttp: cacheNumMisplaced: 0 Jun 29 18:01:04 unNeptunus emhttp: cacheNumExtra: 1 Jun 29 18:01:04 unNeptunus emhttp: shcmd (2463): mkdir -p /mnt/cache Jun 29 18:01:04 unNeptunus emhttp: mount error: No file system (no btrfs UUID) Jun 29 18:01:04 unNeptunus emhttp: shcmd (2464): umount /mnt/cache |& logger Jun 29 18:01:04 unNeptunus root: umount: /mnt/cache: not mounted Jun 29 18:01:04 unNeptunus emhttp: shcmd (2465): rmdir /mnt/cache sdf: Jun 29 18:02:27 unNeptunus emhttp: writing MBR on disk (sdf) with partition 1 offset 64, erased: 0 Jun 29 18:02:28 unNeptunus emhttp: re-reading (sdf) partition table Jun 29 18:02:28 unNeptunus emhttp: shcmd (2620): udevadm settle Jun 29 18:02:28 unNeptunus kernel: sdf: sdf1 Jun 29 18:02:28 unNeptunus kernel: sdf: sdf1 Jun 29 18:02:28 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 18:02:28 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 18:02:28 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 18:02:28 unNeptunus rc.diskinfo[8978]: SIGHUP received, forcing refresh of disks info. Jun 29 18:02:28 unNeptunus rc.diskinfo[8978]: PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo on line 634 Jun 29 18:02:28 unNeptunus rc.diskinfo[8978]: SIGHUP ignored - already refreshing disk info. Jun 29 18:02:28 unNeptunus emhttp: shcmd (2621): set -o pipefail ; mkfs.btrfs -f -K -dsingle -msingle /dev/sdf1 |& logger Jun 29 18:02:29 unNeptunus root: btrfs-progs v4.7.2 Jun 29 18:02:29 unNeptunus root: See http://btrfs.wiki.kernel.org for more information. Jun 29 18:02:29 unNeptunus root: Jun 29 18:02:29 unNeptunus root: Label: (null) Jun 29 18:02:29 unNeptunus root: UUID: 49350932-ba89-4fd9-8fe0-2a81800e7e3c Jun 29 18:02:29 unNeptunus root: Node size: 16384 Jun 29 18:02:29 unNeptunus root: Sector size: 4096 Jun 29 18:02:29 unNeptunus root: Filesystem size: 1.82TiB Jun 29 18:02:29 unNeptunus root: Block group profiles: Jun 29 18:02:29 unNeptunus root: Data: single 8.00MiB Jun 29 18:02:29 unNeptunus root: Metadata: single 8.00MiB Jun 29 18:02:29 unNeptunus root: System: single 4.00MiB Jun 29 18:02:29 unNeptunus root: SSD detected: no Jun 29 18:02:29 unNeptunus root: Incompat features: extref, skinny-metadata Jun 29 18:02:29 unNeptunus root: Number of devices: 1 Jun 29 18:02:29 unNeptunus root: Devices: Jun 29 18:02:29 unNeptunus root: ID SIZE PATH Jun 29 18:02:29 unNeptunus root: 1 1.82TiB /dev/sdf1 Jun 29 18:02:29 unNeptunus root: Jun 29 18:02:29 unNeptunus emhttp: check_pool: /sbin/btrfs filesystem show /dev/sdf1 2>&1 Jun 29 18:02:29 unNeptunus kernel: BTRFS: device fsid 49350932-ba89-4fd9-8fe0-2a81800e7e3c devid 1 transid 5 /dev/sdf1 Jun 29 18:02:29 unNeptunus emhttp: cacheUUID: Jun 29 18:02:29 unNeptunus emhttp: cacheNumDevices: 1 Jun 29 18:02:29 unNeptunus emhttp: cacheTotDevices: 0 Jun 29 18:02:29 unNeptunus emhttp: cacheNumMissing: 0 Jun 29 18:02:29 unNeptunus emhttp: cacheNumMisplaced: 0 Jun 29 18:02:29 unNeptunus emhttp: cacheNumExtra: 1 Jun 29 18:02:29 unNeptunus emhttp: shcmd (2631): mkdir -p /mnt/cache Jun 29 18:02:29 unNeptunus emhttp: mount error: No file system (no btrfs UUID) Jun 29 18:02:29 unNeptunus emhttp: shcmd (2632): umount /mnt/cache |& logger Jun 29 18:02:29 unNeptunus root: umount: /mnt/cache: not mounted Jun 29 18:02:29 unNeptunus emhttp: shcmd (2633): rmdir /mnt/cache Quote Link to comment
JorgeB Posted June 29, 2017 Share Posted June 29, 2017 Sorry, no idea what the problem is. Quote Link to comment
WEHA Posted June 29, 2017 Author Share Posted June 29, 2017 mkay, well thanks for trying Quote Link to comment
WEHA Posted June 29, 2017 Author Share Posted June 29, 2017 (edited) Ok I've been trying some things: Cache slots was set on 2 because I wanted to add 2 drives. When I set cache slots to 1, it works?! Disk is then tagged as normal (green icon) From the moment I set the slots to 2 the first disk is tagged as new (blue icon) and unmountable This with either disk. EDIT: in disk.cfg cacheUUID is still empty, whereas on another test server I have an SSD mounted for cache and that actually shows an UUID Edited June 29, 2017 by WEHA Extra info Quote Link to comment
JorgeB Posted June 30, 2017 Share Posted June 30, 2017 Try this, at least if it doesn't work it may give a better clue of what the problem is: Wipe btrfs from both disks: wipefs -o 0x10040 /dev/sdX You mentioned above that it works with a single device, create a single device btrfs cache and the try to add the second device following method 2 on the FAQ entry below (you can skip the wipe, start on step 2): Paste here any errors Quote Link to comment
WEHA Posted July 1, 2017 Author Share Posted July 1, 2017 Your commands did not do/find anything: wipefs -o 0x10040 /dev/sdg wipefs: /dev/sdg: offset 0x10040 not found wipefs -o 0x10040 /dev/sdf wipefs: /dev/sdf: offset 0x10040 not found One disk was empty so I wiped the other one with the other command: wipefs -a /dev/sdf1 /dev/sdf1: 8 bytes were erased at offset 0x00010040 (btrfs): 5f 42 48 52 66 53 5f 4d wipefs -a /dev/sdf /dev/sdf: 2 bytes were erased at offset 0x000001fe (dos): 55 aa /dev/sdf: calling ioctl to re-read partition table: Success I recreated the cache with only sdg, mounted fine. The next command returned nothing: btrfs device add /dev/sdf1 /mnt/cache I stopped the array and from the moment I selected 2 slots the green dot went to blue, adding the second disk made 2 blue icons Started the array and tried the following command anyway: btrfs balance start --bg --full-balance /mnt/cache Checking status: Every 2.0s: btrfs balance status /mnt/cache ERROR: cannot access '/mnt/cache': No such file or directory btrfs fi show /mnt/cache ERROR: superblock checksum mismatch ERROR: cannot scan /dev/sdc1: Input/output error ERROR: not a valid btrfs filesystem: /mnt/cache sdc is an array disk formatted as btrfs Quote Link to comment
JorgeB Posted July 1, 2017 Share Posted July 1, 2017 10 minutes ago, WEHA said: ERROR: cannot scan /dev/sdc1: Input/output error Your problem may be related to this, this error is not normal and it may be causing issues with any btrfs operation. Quote Link to comment
WEHA Posted July 1, 2017 Author Share Posted July 1, 2017 Ok I will reformat that disk. Am I correct to think that when excluding a disk from the shares and running mover moves the data away from that disk? thanks! Quote Link to comment
JorgeB Posted July 1, 2017 Share Posted July 1, 2017 Looking at your diagnostics sdc is parity, so that explains the error, still it may cause issues, since you have little data move everything from disk1 to another disk and format it xfs just to see if it makes any difference. Quote Link to comment
JorgeB Posted July 1, 2017 Share Posted July 1, 2017 Mover doesn't work like that but you can use midnight commander (mc), make sure you move from disk to disk, eg, /mnt/disk1 to /mnt/disk2 Quote Link to comment
WEHA Posted July 2, 2017 Author Share Posted July 2, 2017 Seems that was the problem, reformatted sdd to xfs (instead of btrfs). Stopped and started the array again and voila, working cache disk. So this is a bug? Quote Link to comment
JorgeB Posted July 2, 2017 Share Posted July 2, 2017 9 hours ago, WEHA said: So this is a bug? Not really, this a consequence of how parity works, the number of btrfs disks and if that number is even or odd, btrfs doesn't like when there's an error on one of the filesystems, and that is likely to cause issues with other btrfs filesystems. Quote Link to comment
Iormangund Posted July 4, 2017 Share Posted July 4, 2017 I see you have this "PHP Warning: Missing argument 2 for force_reload() in /etc/rc.d/rc.diskinfo" in your logs, I am getting that also, will be fine for a while then I get spammed with that in the logs for 10mins then it stops. I don't have an issue with my cache disk afaik so I was curious if this error was related to your issue as no one replying to you has mentioned it and would rather check first before opening a separate thread in case it would be a dupe thread. Quote Link to comment
WEHA Posted July 8, 2017 Author Share Posted July 8, 2017 Sorry for the late reply but I wasn't able to turn off and on the array. I don't have the errors anymore so I suppose it must be related somehow to a bad btrfs. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.