[Support] MariaDB Official

[Pete0]

Today I was offered an update to the MariaDB container and installed, which I normally always do.

 

Checking the protocol after starting I saw this messages:

2024-02-14 18:46:48+01:00 [Note] [Entrypoint]: Entrypoint script for MariaDB Server 1:11.2.3+maria~ubu2204 started.
2024-02-14 18:46:48+01:00 [Note] [Entrypoint]: MariaDB heathcheck configation file missing, assuming desirable
2024-02-14 18:46:48+01:00 [Note] [Entrypoint]: MariaDB upgrade (mariadb-upgrade or creating healthcheck users) required, but skipped due to $MARIADB_AUTO_UPGRADE setting

 

I know of the need sometimes to upgrade mariadb after an update but this time the message contains something of a healthcheck.

 

Yet I still tryed to upgrade the database which was not required according to the output of mariadb-upgrade.

 

I found out that the container has a shell script healthcheck.sh included but I don't know how to run it properly - or what exactly to do to get rid of the message.

 

EDIT:

I just added the environment variable MARIADB_AUTO_UPGRADE to the container - because I always intended to in the past so I wouldn't have to manually upgrade when necessery. After restarting the contianer with the variable the message was gone.

 

Not sure why the manual upgrade didn't fix the issue but setting the variable seems to have done so. Maybe there is a parameter for mariadb-upgrade that I didn't use.

 

2024-02-14 18:58:21+01:00 [Note] [Entrypoint]: Entrypoint script for MariaDB Server 1:11.2.3+maria~ubu2204 started.
2024-02-14 18:58:21+01:00 [Note] [Entrypoint]: MariaDB heathcheck configation file missing, assuming desirable
2024-02-14 18:58:21+01:00 [Note] [Entrypoint]: Starting temporary server
2024-02-14 18:58:21+01:00 [Note] [Entrypoint]: Waiting for server startup
2024-02-14 18:58:22+01:00 [Note] [Entrypoint]: Temporary server started.
2024-02-14 18:58:22+01:00 [Note] [Entrypoint]: Backing up system database to system_mysql_backup_11.2.3-MariaDB.sql.zst
2024-02-14 18:58:22+01:00 [Note] [Entrypoint]: Backing up complete
2024-02-14 18:58:22+01:00 [Note] [Entrypoint]: Creating healthcheck users
2024-02-14 18:58:32+01:00 [Note] [Entrypoint]: Entrypoint script for MariaDB Server 1:11.2.3+maria~ubu2204 started.
2024-02-14 18:58:32+01:00 [Note] [Entrypoint]: MariaDB upgrade not required

 

Edited February 14 by Pete0

[Autchirion]

I just updated to the version 11.3.2 which was released 7h ago, since then mariadb is broken. I had to roll back to a backup from tonight and go back to version mariadb:11.2.3.

 

Anyone else facing this issue? I wasn't able to connect any more, every time phpmyadmin/own scripts/nextcloud tried to connect it would resolve in an errormessage posted in the log.

Edited February 21 by Autchirion

[juan2023]

I have the same problems.

1) The first one was solved adding the MARIADB_AUTO_UPGRADE variable (1).

2) No application can connect to the ddbb. I have several data bases, but all appplications get the same error:

From phpmyadmin: "mysqli::real_connect(): (HY000/2054): Server sent charset (0) unknown to the client."

Also very similar from Nextcloud: "SQLSTATE[HY000] [2054] Server sent charset (0) unknown to the client."

...

The log of MariaDB just get the same entry for every connection attemtp:

"[Warning] Aborted connection 270 to db: 'unconnected' user: 'unauthenticated' host: '172.23.0.1' (This connection closed normally without authentication)"

 

Anyone has been able to solve this connectivity errors??

 

EDIT:

I have installed a previous version as suggested by Autchirion and it is working fine.

At least Nextcloud and phpmyadmin, I have to test Home Assistant.

I have used REpository: mariadb:11.2.3 as Autchirion already mentioned.

 

Edited February 21 by juan2023

[RichardU]

I have two Wordpress installations and both were dead this morning. Editing the MariaDB docker and switching the Repository to mariadb:11.2.3 fixed it. I'm glad there seems to be no data loss.

[serbanm33]

1 hour ago, RichardU said:

I have two Wordpress installations and both were dead this morning. Editing the MariaDB docker and switching the Repository to mariadb:11.2.3 fixed it. I'm glad there seems to be no data loss.

 

Thank you very much, changing the repository to mariadb:11.2.3 fixed the issue for me

[projektdotnet]

I also had to roll-back to 11.2.3 which did fix my issue (the error was about unknown charset (0))

[Team-O]

Same problem here. My Firefly III container was unable to connect to the db (Server sent charset unknown to the client.). Downgrade to 11.2.3 fixed it.

[Gvon]

just add this to Post Arguments: --transaction-isolation=READ-COMMITTED --log-bin=binlog --binlog-format=ROW --character-set-server=utf8

 

https://github.com/nextcloud/docker/issues/2165

[Gvon]

update MariaDB 11.4.0 up seem fix the problem no need to add this to Post Arguments: --transaction-isolation=READ-COMMITTED --log-bin=binlog --binlog-format=ROW --character-set-server=utf8

[Hepheastus]

On 2/21/2024 at 6:33 AM, Autchirion said:

I just updated to the version 11.3.2 which was released 7h ago, since then mariadb is broken. I had to roll back to a backup from tonight and go back to version mariadb:11.2.3.

 

Anyone else facing this issue? I wasn't able to connect any more, every time phpmyadmin/own scripts/nextcloud tried to connect it would resolve in an errormessage posted in the log.

I had spent four hours or more with little sleep trying to figure out how to allow access from WordPress user through 172 IP group to Maria, which was a problem as I host a shirt store for myself and others until I finally stumbled onto the support button and found this fix, it really was a savior.  I hadn't connected the dots it was an update issue, I thought it was an issue of a failed database.

 

11 hours ago, Gvon said:

update MariaDB 11.4.0 up seem fix the problem no need to add this to Post Arguments: --transaction-isolation=READ-COMMITTED --log-bin=binlog --binlog-format=ROW --character-set-server=utf8

I put my repository back to default, had a database failure message on the site again, rechecked for update and had one available.

After update, the site was back to working function so this was also a fix.

[Gvon]

8 hours ago, Hepheastus said:

I put my repository back to default, had a database failure message on the site again, rechecked for update and had one available.

After update, the site was back to working function so this was also a fix.

Yes you updated on latest version that fixed the problems 

[-C-]

On 2/22/2024 at 5:26 AM, Gvon said:

update MariaDB 11.4.0 up seem fix the problem no need to add this to Post Arguments: --transaction-isolation=READ-COMMITTED --log-bin=binlog --binlog-format=ROW --character-set-server=utf8

 

I've checked for updates on the Docker page, it's saying I'm on the latest version, but when I check the log I seem to be stuck on 11.3.2:
 

2024-02-25 15:34:48 0 [Warning] Can't create test file '/var/lib/mysql/37ddd26137c4.lower-test' (Errcode: 13 "Permission denied")
2024-02-25 15:34:48 0 [Note] Starting MariaDB 11.3.2-MariaDB-1:11.3.2+maria~ubu2204-log source revision 068a6819eb63bcb01fdfa037c9bf3bf63c33ee42 as process 1

 

How do I get 11.4.0?
Thanks

 

Edit: I've tried added the post arguments with 11.3.2and still getting the errors.

Edited February 25 by -C-

[djgrazzy]

On 2/26/2024 at 2:42 AM, -C- said:

How do I get 11.4.0?
Thanks

Modify the Repository to something like this.

mariadb:11.4.1-rc

[-C-]

I've now tried upgrading to 11.4.1-rc and downgrading to 11.2.3 and it's still refusing to start.

 

Seems to be falling over on startup with the same "Can't create test file" error.

 

Here's the complete log:

 

2024-02-27 20:37:36 0 [Warning] Can't create test file '/var/lib/mysql/1dd699d8d500.lower-test' (Errcode: 13 "Permission denied")
2024-02-27 20:37:36 0 [Note] Starting MariaDB 11.4.1-MariaDB-1:11.4.1+maria~ubu2204-log source revision fa69b085b10f19a3a8b6e7adab27c104924333ae as process 28
2024-02-27 20:37:36 0 [ERROR] mariadbd: File './binlog.index' not found (Errcode: 13 "Permission denied")
2024-02-27 20:37:36 0 [ERROR] Aborting
2024-02-27 20:38:09+00:00 [ERROR] [Entrypoint]: Unable to start server.
2024-02-27 20:37:27+00:00 [Note] [Entrypoint]: Entrypoint script for MariaDB Server 1:11.4.1+maria~ubu2204 started.
2024-02-27 20:37:35+00:00 [Note] [Entrypoint]: MariaDB upgrade information missing, assuming required
2024-02-27 20:37:35+00:00 [Note] [Entrypoint]: Starting temporary server
2024-02-27 20:37:35+00:00 [Note] [Entrypoint]: Waiting for server startup

 

I've got multiple daily appdata backups going back months- from both the appdata backup plugin, and ZFS snapshots.

 

I've tried reverting to a snapshot from before the problem, but it's not making any difference.

 

I've never had to repair/ restore a broken docker app before. Is there something I'm missing?

 

Could the snapshot restoration not be working because I've changed the versions in the docker script which is causing a mismatch?

 

Would really appreciate some help with getting this back up as I have a few other things (mostly Nextcloud) that are depending on it.

 

Thanks

[djgrazzy]

8 hours ago, -C- said:

 

Could the snapshot restoration not be working because I've changed the versions in the docker script which is causing a mismatch?

 

I don't believe so no. It appears to be a permissions issue.

This command should show your current permissions (where 'mariadb'  is your mapped volume name)
 

ls -la /mnt/user/appdata/mariadb

and should result in something like this 
............................................................................................................
# ls -la /mnt/user/appdata/mariadb/
total 0
drwxr-xr-x 1 root   root    20 Mar  9  2023 ./
drwxrwxrwx 1 nobody users 1194 Feb 13 21:47 ../
drwxr-xr-x 1 nobody users    0 Mar  9  2023 config/
drwxr-xr-x 1 nobody users  676 Feb 28 05:13 data/

............................................................................................................

and can then be fixed with something like this. someone pls correct me if I'm wrong.
 

chmod -R 777 /mnt/user/appdata/mariadb

 

[-C-]

Thanks-

 

All files under /mnt/user/appdata/mariadb-official are now like this:

-rwxrwxrwx 1 999 users    8192 Jan 23  2023 column_stats.MAD*

 

Mariadb's running, but Nextcloud still isn't working. I'm getting this in the Mariadb log:

2024-02-28 22:49:12 14 [Warning] Aborted connection 14 to db: 'nextcloud' user: 'nextcloud' host: '172.17.0.1' (Got an error reading communication packets)

 

I then tried logging in with Adminer, which got this error in Adminer and the same in the Mariadb log:

 

Access denied for user 'root'@'172.17.0.1' (using password: YES)

 

My understanding was that a ZFS snapshot would put everything back exactly as it was when the snapshot was taken. How could the permissions be different, and why would they have changed anyway when I haven't been anywhere near that directory?

 

Any ideas?

[djgrazzy]

I can't comment on the snapshots sorry. You mentioned you have appdata backup plugin backups as well, can you try those?

[Gvon]

On 2/28/2024 at 4:04 AM, -C- said:

Restore your snapshot first then change to this repository "mariadb:latest" after that click check for update you will see new version available to update 

 

Edited February 29 by Gvon

[-C-]

14 hours ago, Gvon said:

Restore your snapshot first then change to this repository "mariadb:latest" after that click check for update you will see new version available to update 

Unfortunately, that hasn't worked- rolled back to an earlier snapshot, then did as you said but didn't get a new version when I switched to :latest

Everything looks to start up fine in the Mariadb log, but still not able to log in (tried console & Adminer).

[Gvon]

19 minutes ago, -C- said:

Everything looks to start up fine in the Mariadb log, but still not able to log in (tried console & Adminer).

Did you set Generate Random Password to yes ? If so check your password in the mariadb log overy time you restart the password will change 

Edited March 1 by Gvon

[-C-]

23 hours ago, Gvon said:

Did you set Generate Random Password to yes ? If so check your password in the mariadb log overy time you restart the password will change 

Thanks- I'm not seeing the password on restart, but I was able to log in with the nextcloud acct rather than root.

I'll figure out the root password another time.
 

[ShuaiTianShi]

Good evening everybody,

 

I got some issues with my database and photoprism.

After updating my unraid to Version 6.12.10 and all my docker Containers, my maria-db Container will not start.

 

I bot the following message, but i'm nee to unraid and the complex subject docker Container

 

 

What should I do now?

What command do I have to make and where?

 

Thanks for your help.

 

Greetings

unraid-diagnostics-20240410-2049.zip