Mirror2.freepbx.org and mirror1.freepbx.org Down

Went to go do a module update and I am getting a connection error to both mirror2.freepbx.org and mirror1.freepbx.org.

It appears these are down at the moment. AS well as LE is timing out with the following ‘There was an error updating the certificate: Connection timed out after 10001 milliseconds’ even though ALL ports both in and out are opened up to the FreePBX server.

It’s sad how often the FreePBX mirrors are down. This is a fairly routine occurrence on weekends. Most of us would have lost our jobs by now if we couldn’t do better.

The certman module being dependent on the mirror for LetsEncrypt is pointless and an all around bad design decision. It does more harm than good even when the mirrors are up.

The https://mirror.freepbx.org repo is working fine; just tested on my own servers.

Unfortunately changing the repo doesn’t do anything for certman.

There are plenty of ways to work around the certificate updater, including waiting until later if it’s just a renewal, or grabbing one manually using a couple quick commands at the command line.

ps: I don’t disagree that the repo dependency ought to be written out of certman.

Same here on multiple boxes across the US.

The funny thing to me is that they aren’t mirrors at all; they are the original repos (at least as far as the public is concerned). No code or method has been released to allow people to create an actual “mirror.”

1 Like

“Mirrors” are usually meant to give some measure of redundancy. In the FreePBX case, it seems they are always down together.

I’m pretty sure if someone went through added it all up there would days of unplanned downtime per year, not minutes or hours.

And apparently no effective monitoring…

what’s most frustrating about this is it’s the weekends where I need to do the updates and upgrades and need the servers the most. Can’t mess with a phone system during business hours.

Looks like back up - working now on at least 3 of my pbxes.

Hey guys,

We updated some of the mirror server infrastructure a while ago and it’s been running through the weekend update flood fine for a bit, but for some reason this weekend it had some trouble again.

We’re going to try to spend some time isolating better what might be causing it to get overloaded.

For the time being, one thing that you can do is set your automatic update time for your system to be something other than friday night - perhaps saturday or sunday night instead, as friday nights are peak traffic.

Sorry again about the trouble.

Matthew Fredrickson

1 Like

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.