Community.openenergymonitor.org temporarily offline

Organisation Forum
1

Sorry, this is off-topic for the thread, for reasons explained below,
so perhaps somebody could move it to its own thread?

I can’t access
https://community.openenergymonitor.org/t/radio-interference-xmas/12736/12
or indeed the community base page. I can access eg. the shop.

I get an Error 502 and a banner at the top says:

This page
(Radio interference - Xmas? - #12)
is currently offline. However, because the site uses Cloudflare’s
Always Online™ technology you can continue to surf a snapshot of the
site. We will keep checking in the background and, as soon as the site
comes back, you will automatically be served the live version.

If I do try to access the snapshot, I get apparently the same error
page (or very similar - I haven’t read it in detail) but without the
suggestion to access a snapshot. AIUI that isn’t supposed to happen. Is
there a problem with Cloudflare?

Full disclosure: A few minutes ago I was having a major problem with
Firefox 68.2.0esr (64-bit) on my openSUSE Leap 15.0 system. It wasn’t
able to access any websites at all. I use DoH and had set it up a while
ago; apparently something just changed to break it and by turning it
off and back on again using the new switch in preferences everything
started working again. This may be relevant because the DNS provider is
Cloudflare, so maybe it’s them that has changed something and broken
various things?

Cheers, Dave

2

Glyn says there was a problem with an update.

3

Thanks, Robert. Glyn/Megni or Cloudflare update?

4

That’s all I know.

5

Ah, thanks again. @glyn.hudson could you please clarify? I’m interested whether it might explain some other problems elsewhere.

6

It was down on my end too. (US)

7

Yes the forum was down earlier today, Glyn was performing a security update and something went wrong, all sorted now I believe.

8

Well, like they say, go big, or go home! :wink: :grin:

1 Like
9

It was a bug in the latest version of Discouse, it was fixed by the developers by reverting the commit:

https://github.com/discourse/discourse/commit/7b6bafc651b27094bfd05f46049e7e468208d659

It effected a number of installations including ours:

There was no change of any data loss since the forum data is stored outside the docker container and backup to several locations each night.

Very sorry for the hassle.

1 Like
10

Thanks for the update @glyn.hudson Wasn’t a problem for me. I just had a problem with FF DoH at the same time and since that also involves Cloudflare, I put 2 and 2 together and got 5 :grin:

11

No issues, at all with Cloudflare. The error message which was displayed was Cloudflare unable to connect to our Discourse server and trying to server a cached version.

12

Always interesting to learn the reason.

13

Yes, it was its inability to find a cached version which I found interesting. The whole point of that cache is to keep the site up (at least for viewing) if the original can’t be contacted, AIUI.

14

It’s because Discourse is a web app rather than static pages, couldflare is able to cache static pages put not a dynamic app. However, there is still value in using cloudflare, it caches gui elements, css and provides protection from ddos attacks.

1 Like