Unable to login and to access admin-page after update(s)

Hello

Last week I updated both my self-hosted Bitwarden instance and the MariaDB instance on my Docker setup on Synology which I had installed via Portainer Stack following these instructions:

The Docker images involved are as follows:
• bitwarden/self-host:beta
• mariadb:11.4-noble

Since the update, after entering my password, I receive an error message stating that an unexpected error has occurred. I can no longer log in, and in particular, cannot back up the organization’s data. Additionally, my attempt to access the admin page has been failing since then (Bad Gateway).

The bitwarden container logs show the following:

2024-10-31 08:38:37,103 INFO spawned: ‘admin’ with pid 1011
2024-10-31 08:38:52,124 INFO success: admin entered RUNNING state, process has stayed up for > than 15 seconds (startsecs)
2024-10-31 08:39:05,162 WARN exited: admin (terminated by SIGABRT (core dumped); not expected)
2024-10-31 08:39:06,164 INFO spawned: ‘admin’ with pid 1043
2024-10-31 08:39:21,187 INFO success: admin entered RUNNING state, process has stayed up for > than 15 seconds (startsecs)

The mariadb container logs show the following:

2024-10-31 9:25:35 5 [Warning] Aborted connection 5 to db: ‘bitwarden_vault’ user: ‘bitwardenuser’ host: ‘’ (Got an error reading communication packets)
2024-10-31 9:26:04 9 [Warning] Aborted connection 9 to db: ‘bitwarden_vault’ user: ‘bitwardenuser’ host: ‘’ (Got an error reading communication packets)

In the meantime, I have deleted the containers, backed up the existing Docker folders to a different location, and then deleted the docker folders to perform a fresh setup. However, even after such a clean reinstallation(s), the same errors persist. I am able to register, but I cannot log in afterward and can’t access the admin page.

I updated both containers several times in the past without any issues and there was nothing I did in another way now.

The official support didn‘ react at first and told me the following on Monday:

„May i ask that you try this;

This issue could be happening because of permissions when the Docker Synology application is enabled on a volume other than the default volume created in Synology. (for example, /volume6) when the Docker Synology app is installed, two directories are created: /volume6/docker and /volume6/@docker.

However, the permissions are different between the two; the former directory gets write access for the docker_bitwarden user, while the latter does not. Changing the permissions on the latter:

chown -R docker_bitwarden @docker/“

Unfortunately this didn’t help. I immediately informed the support and since then – again – they do not react.

Can anyone please help since Bitwarden, including the paid license, is currently of no use to me?

Thank you…

1 Like

I have exactly the same issue.
But what I have noticed is that I still have access with the edge extension
I can’t log locally or using the dyndns url to the web page and I can’t access to the admin (nginx error).
I checked portainer logs and I have exactly the same problem.

Btw I just did a pull update for the stack (bitwarden/self-host:beta).

To resume:

  • Web client Acces: Nok
  • Edge bitwarden extension access: Working
  • iOS/Android Bitwarden app: Nok
  • Web admin access: NGINX 502 badgateway

This is a known issue unfortunately, and from what I’ve read on GitHub it’s not getting attention (that we know of) from Bitwarden. Two things to add here. First, if you have a device that still has Bitwarden access to your account, export your passwords. Second, I (and many others) had the same problem, and just going back to self-host:2024.9.2-beta allowed me to log in again. None of the 10.x images has worked for me so far, but the 9.x still do. Good luck!

1 Like

Thank you. It is good to hear that I am not the only one facing the problems. On the other hand it is disappointing to hear that this is a longer known issue and BW is obviously not able to fix it.

There must be a reason for the 10.x images. I hope there is no security issue using a 9.x image…

I don’t think that it’s they aren’t able, it just that since this is beta it’s not going to be on the top of their to-do list.

I saw you edited your post, but to answer the original question, mariadb version is 11.3-jammy. Perhaps a fresh install of 9.2, not a clone of the 10.x and image update, might do it? I too use Portainer and always keep my old containers (stopped) until the new one is confirmed working. Good luck, not having access to your passwords and such is awful.

There is. The month changed. Bitwarden uses date-based versioning. 2024.10.x indicates the release process started in the 10th month (October) of 2024.

The release notes can help you decide if you are comfortable using an older release.

Would be helpful for this discussion if you could link the open GitHub issue that you are referring to.

Issue opened on GitHub 3 weeks ago, exact same results as OP, I should have posted this earlier:

Thank you. I edited it because finalyy I got it to work with 9.2 (with mariadb:11.4-noble). That way I can at least also access my organisation again…

Thank you - I will have a look into the notes…

Bitwarden Support announced an update on Nov. 14 which they say will fix the problem…

2 Likes

That’s great news, can you provide a link or reference?

It was an email which says:

…I believe we’ve identified the cause of this error, and the fix has been implemented here:

Update 2024-09-05_00_SyncDuoVersionFourMetadataToVersionTwo.sql (#4982) · bitwarden/server@639ee57 · GitHub


This should be resolved in the next update, which is planned to be released next Thursday, the 14th.

Please let us know if this resolves the issue for you, or if you still see errors after applying the update.“

Just an FYI for anyone with this admin page access issue, I installed the latest Docker beta version and it’s working again!

1 Like

Thank you, Alphanaut - I also updated it, but when I open the web vault it says

“© 2024 Bitwarden Inc. 2024.9.1”

at the bottom. I tried again by repulling the beta-image but it is the same.

Would’t “2024.11.1” be the correct details of the version?

I know that when I updated to the latest self-host release, a lot of times the web interface version number lags behind. Then at some point in a future update it catches up. Not sure the reason for that, but I’ve seen it many times. Even now the Admin interface is showing I’m slightly behind after the most recent update (below).