Problem custom DNS after upgrading in MIAB 0.53

[M1k3]

Hi,

I had configured several custom DNS with version 0.52.

After upgrading MIAB to 0.53, my Custom DNS disappeared, and when I go to the “Custom DNS” page I have an error message that appears “Something went wrong, sorry”.

I have the same error message when I want to recreate my custom DNS.

Thank you for your help

[openletter]

Please post this as an issue to the project GitHub:

mail-in-a-box/mailinabox

[M1k3]

After some research, I found, I think, the solution (finally a method to solve my problem). I deleted the file /home/user-data/dns/custom.yaml, to “reset” the custom DNS, and then I recreated manually each custom DNS (I didn’t have many of them fortunately).

[mveplus]

Hi [M1k3],

Thanks for your question! And yes, backing up and deleting custom.yaml get’s rid of the error, but then have to recreate all of the records, if you have a few custom records that’s okay but if you have multiple domains with multiple custom records you do not really want to do this by hand, but rather this needs investigation and fix on the backend.

I have updated today and v.0.52 to v.0.53 and sow it by coincidence.
Linking the Github issue for anyone that has the same issue and can provide further info.

N.B. the fix turn to be quite simple checkout the GitHub link above.

Cheers,
Martin

[admecoach]

I have a large number of custom records I can’t access anymore from the admin based on this issue. Wondering if anyone else is having the error with their Custom DNS tab after the v0.53 update or a fix other than backing up and deleting the custom.yaml file?

[blinkingline]

You might consider reviewing the custom.yaml first to make sure there’s not a domain in there which is no longer needed. You should be OK to edit custom.yaml to remove it directly.

Otherwise, until the issue gets resolved run dns_update.py from the server’s command line to see what is in the Custom DNS setup, then use that to generate how your custom config is set up, delete the custom.yaml and regenerate it.

[alento]

Just to clarify here … as I fixed an install yesterday with this exact issue.

Manually review and edit the custom.yaml file and be certain that any entries pertaining to domains which no longer have a valid email address are gone.

Then run the dns_update.py script after editing custom.yaml.

Once this is done, all should function normally.

[admecoach]

Confirming fix after locating DNS for domain entries no longer on this instance still present in the custom.yaml

sudo nano /home/user-data/dns/custom.yaml
killed 2 lines of older domain dns

ran
sudo ./mailinabox/tools/dns_update --force

Custom DNS correctly presents from web admin.

Thanks @alento & others guiding fix here and on github. I think i’ve archived some email accounts/domains without always cleaning up custom dns first.