[Bug] Maxed out DNS Requests with no Backoff or Retries
-
Hey Cloudron team!
I just migrated my Cloudron to a new server and had some issues around updating DNS entries to the new IP address. Specifically I use Namecheap, but either way, from the logs below we can see there is logic that can probably be optimized. I have about 40 apps or so installed on my box, so if a DNS provider has any rate limiting on their API, it can cause issues when doing a complete DNS migration/update. See log below for what I was seeing:
Jan 06 11:21:41box:dns registerLocation: Get error. retryable: true. Too many requests
In this case, I would expect Cloudron to gracefully register this error (assuming we're getting a legitimate 429 status code from the DNS API), and retry with exponential backoffs until all queries are completed. Instead, this failed on all 3 of my domains since the first one has a lot of apps, and therefore DNS entries, to complete successfully.
A few related thoughts:
- I love Cloudron! The migration as a whole was pretty smooth even given this DNS hiccup. I just went into my Namecheap dashboard and updated them all manually not ideal, but not the end of the world. Besides that the upgrade process went without a hitch. Thanks for all that you do to help make it so!
- I was hoping to register this bug in your Cloudron Gitlab instance but I don't think I have permissions for that, even when I was logged in? I hope this is the right place to report this issue.
- The reason I moved servers is due to Cloudron 7.6.2 needing AVX support. I personally think that such a large dependency should have been introduced in a major version (ie 8.0.0) so that users like myself wouldn't be 'stranded' on a random version like 7.6.1 and not able to receive any updates at all (including on the apps, because of an issue reported in a separate thread on the forum here).
-
-
-