Having Issues with my NAS. Shares missing, WebUI broken.

[Jocko]

Hi

I deepened more your issue.

Sorry but I can't quite figure out what you mean. Do you care to clarify?

If you do not understand me that means from the console, you did not ask to "resend the request" or "open the URl in a new tab" with the context menu on the url line.

So to get the response "Invalid request" that means you have enabled the option "no referrer" on the server apache or by another way you locked referrer on FF. Maybe you used this trick to change the default value of network.http.sendRefererHeader : https://stackoverrun.com/fr/q/2280541. I got the same behaviour as yours with this change :

no-referrer.jpg

If you did not change it then I advice to reinstall the current fw if the web-interface still works on Edge : so open the menu "Firmware update" and click on the button "Update firmware via fvdw-sl server (latest version)" even if the menu claims you have the latest version

[Woodani]

Hi,

Since we last spoke about this issue I have actually gotten an new PC. The issue dose not exist on my new PC at all. The webui loads correctly in all my browsers. It seems the issue was isolated to just my old PC.

I did not make any of the changes you had mentioned. I still have my old PC if you want to keep troubleshooting this for you're own understanding we can. But as far as I am concerned my issue is resolved.

On a side note I just started the process to change my chunk size. Hopefully it wont cause any issues.

Thanks.

[Jocko]

Hi

Glad to heard this.
Anyhow something has changed your Firefox config. Yes, for my understanding (and it will be the last action), can you follow this guidance (only the FF step and not the malware check) to restore the default FF config and check if the webgui is loaded this time.

https://malwaretips.com/blogs/reset-firefox-settings/

[Woodani]

Hey,

I just reset firefox per your instructions and that fixed the issue. Firefox loaded the webui without issues. I checked with chrome and the issue remained so then I did the "Restore settings to their original defaults" in chrome and that fixed chrome as well.

I wonder what browser settings where screwing up both chrome and firefox? :thinking Possibly a browser extension?

On another subject, when I attempted to change my chunk size with the command you gave me it failed with an error about a missing backup. This is the output of the terminal.

Code: Select all

root@fvdwsl-base:/ # mdadm --grow --chunk=128 /dev/md0
mdadm: /dev/md0: Cannot grow - need backup-file
mdadm:  Please provide one with "--backup=..."

Any idea how to resolve that issue?

Thanks again.

[Jocko]

Hi

I wonder what browser settings where screwing up both chrome and firefox? :thinking Possibly a browser extension?

I don't think so. Chrome uses a strong cache and the common ways to force the browser to update its cache do not work some time with Chrome...

I think your issue was the browser cache with chrome.

On another subject, when I attempted to change my chunk size with the command you gave me it failed with an error about a missing backup. This is the output of the terminal.

Why do you think you have to change again the chunk size on your raid ? When you do it you do not have to repeat it later. (and I think it is why mdadm asked you a backup file for storing some critical data.

To check that, please to post

Code: Select all

cat /proc/mdstat

[Woodani]

It never worked, it failed with that error each time I tried it. I just now remembered to post that here and ask about the error. If I'm not mistaken this shows that the chunk is still 512k so it never resized.

Code: Select all

root@fvdwsl-base:/ # cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md0 : active raid5 sda8[0] sde8[4] sdd8[3] sdc8[2] sdb8[1]
      15616464896 blocks super 1.0 level 5, 512k chunk, algorithm 2 [5/5] [UUUUU]

unused devices: <none>

[Jocko]

Hi

about chunk yes you can change it without destroying the raid.

If you want to change the value, run this command

Code: Select all

mdadm --grow --chunk=128 /dev/md0

But you should run it only when the check is complete and note this action takes more time than a check job (here around 4 days)

I did not notice you failed to change the chunk size on your raid

As it seems mdadm want a backup file to perform it on your raid,the new command is

Code: Select all

mdadm --grow --chunk=128 /dev/md0 --backup-file=/rw_fs/md0-chunk-backup

Please to keep this location (backup not destroyed on nas rebooting and stored in a partition with enough free size)

Before running it check if your raid status is clean

root@Acrab:/ # mdadm --detail /dev/md0
/dev/md0:
Version : 1.0
Creation Time : Wed Dec 3 14:40:52 2014
Raid Level : raid5
Array Size : 11709457408 (11167.01 GiB 11990.48 GB)
Used Dev Size : 2927364352 (2791.75 GiB 2997.62 GB)
Raid Devices : 5
Total Devices : 5
Persistence : Superblock is persistent

Update Time : Sat Jan 30 10:15:17 2021
State : clean <-------------------- here
Active Devices : 5

Additional information,
- several days will be required to complete the action. So do not shut down your NAS while the chunk resizing action is in progress (to check it

Code: Select all

cat /proc/mdstat

- mdadm may also create a bitmap cache. So do not remove it and when the action will complete you may run the command

Code: Select all

mdadm --grow --bitmap=none /dev/md0

- I found this bitmap cache may also be created on a resync action when you have a degraded raid. I think it was the case previously with your raid
- about chunk resize, for your understanding:

Please note that changing the chunk size is a very slow process because this involves rewriting all data that is stored in the RAID array. As this is done in a way that ensures data safety even in case of a system crash, each piece of data actually has to be written twice. Therefore, you should expect this process to take quite a while (in the order of days).

[Woodani]

Hi,

Something strange is happening. I ran the detail command and verified that my raid status was clean, Then I ran the new grow command as you posted, with the backup specified. It again failed saying no such file or directory. Before I could copy the error to post here my SSH connection crashed. My entire terminal crashed completely.

At this point I SSHed back into the NAS and was going to run the two commands again to show you the output and the error. When I ran the first command to check the raid status I noticed it is showing something different.

Code: Select all

root@fvdwsl-base:/ # mdadm --detail /dev/md0
/dev/md0:
        Version : 1.0
  Creation Time : Mon Jan  1 07:03:58 2018
     Raid Level : raid5
     Array Size : 15616464896 (14893.02 GiB 15991.26 GB)
  Used Dev Size : 3904116224 (3723.26 GiB 3997.82 GB)
   Raid Devices : 5
  Total Devices : 5
    Persistence : Superblock is persistent

    Update Time : Sat Jan 30 11:38:18 2021
          State : clean, reshaping
 Active Devices : 5
Working Devices : 5
 Failed Devices : 0
  Spare Devices : 0

         Layout : left-symmetric
     Chunk Size : 512K

 Reshape Status : 0% complete
  New Chunksize : 128K

           Name : fvdwsl-base.local:0  (local to host fvdwsl-base.local)
           UUID : 35290b17:a118a68b:8ca0e64f:f1a01f89
         Events : 191

    Number   Major   Minor   RaidDevice State
       0       8       40        0      active sync   /dev/sda8
       1       8       24        1      active sync   /dev/sdb8
       2       8        8        2      active sync   /dev/sdc8
       3       8       72        3      active sync   /dev/sdd8
       4       8       56        4      active sync   /dev/sde8

It appears that even though it said it failed with the "unknown file or directory" error it is still resizing the chunk anyway.

The cat command you posted seems to confirm that.

Code: Select all

cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md0 : active raid5 sda8[0] sde8[4] sdd8[3] sdc8[2] sdb8[1]
      15616464896 blocks super 1.0 level 5, 128k chunk, algorithm 2 [5/5] [UUUUU]
      [>....................]  reshape =  0.0% (1097728/3904116224) finish=22297.8min speed=2916K/sec

unused devices: <none>

I'm assuming I should just leave it alone and let it run until it finishes now?

Thanks again for all your help.

[Jocko]

Yes amazing but reshaping seems to work as expected and yes let it run until complete

Just for my understanding (because it is not a common use command), please to post

Code: Select all

ls -l /rw_fs/md0-chunk-backup

to check if the file is really present

[Woodani]

Ok here is the output of that.

Code: Select all

root@fvdwsl-base:/ # ls -l /rw_fs/md0-chunk-backup
-rw-------  1 root root 33558528 2021-01-30 16:09 /rw_fs/md0-chunk-backup

It looks like it is present but it is timestamped as the exact time I ran the command. Does that mean it was just created or am I looking at the wrong?