Was there a point to the forum being completely down for 24 hours? (1 Viewer)

[Lord Dalek]

...or just the usual "Crowdflare crashed"?

 

[dpippel]

I'd be interested in knowing what happened as well. Also, with the Forum down for such a long time yesterday, why was there no announcement about the outage on the Home Theater Forum Facebook page?

@Ronald Epstein, is there anything you can share?

 

[Robert Crawford]

One of the owners will probably say more about what happened but we had some serious server issues.

 

[Edwin-S]

I figured they must have had a massive crash to be offline for that long.

 

[Edwin-S]

I figured they must have had a massive crash to be offline for that long

I'd be interested in knowing what happened as well. Also, with the Forum down for such a long time yesterday, why was there no announcement about the outage on the Home Theater Forum Facebook page?

@Ronald Epstein, is there anything you can share?

Their whole server had to be down to be directed to the wayback machine, so I doubt they could have posted a message; otherwise, I'm sure they would have.

 

[Jonathan]

Yes, I want to shed some light on the unexpected 22-hour downtime yesterday. A solid-state drive in our server failed in the middle of the night. It took a while for our hosting provider to track down the issue. After figuring out the cause, it was quickly replaced, but getting all the data back from the last backup on Oct 3rd at 11pm Pacific took a whopping 14 hours. Sadly, we lost a 5-hour window of data right before the crash. I'm really sorry for any hassle this caused.

We're back in action now, and I'm looking into ways to prevent this from happening again.

Thanks for hanging in there with us.

 

[JohnRice]

It might be my imagination, but it seemed like the forum was running slow in the hours leading up to the failure.

 

[Carlo_M]

Thanks for the transparency, Jonathan! While hardware failure is unfortunate, it's much better than what I was fearing (e.g. a security breach + DDoS attack) so while it's a bummer that we were down for a bit, it could have been a much worse!

 

[Jeffrey D]

Don’t let it happen again!

Glad we’re back up.

 

[BobO'Link]

I'm more curious as to why a hosting service isn't using a hypercore type system which pretty much negates a *single* drive failure taking down a site. At the very least running on a RAID configuration that would take such failure into account and allow services to continue while a replacement drive is installed and the RAID rebuilt.

 

[Nelson Au]

It might be my imagination, but it seemed like the forum was running slow in the hours leading up to the failure.

I noticed that too! I thought maybe it was because my ISP wasn’t running at full speed as I’ve been experiencing slow downs.

Thanks Jonathan for the explanation for the service interruption.

 

[dpippel]

I figured they must have had a massive crash to be offline for that long

Their whole server had to be down to be directed to the wayback machine, so I doubt they could have posted a message; otherwise, I'm sure they would have.

I think you misunderstood what I said. They could have posted a message about the outage on the HTF Facebook page. They did not.

 

[Edwin-S]

I think you misunderstood what I said. They could have posted a message about the outage on the HTF Facebook page. They did not.

Hmm, yeah. I never thought of that. I don't have Facebook so it never even crossed my mind that they could have posted an announcement there.

 

[DaveF]

Hmm, yeah. I never thought of that. I don't have Facebook so it never even crossed my mind that they could have posted an announcement there.

I can get you Facebook. Cheap. I know a guy. LMK. I’ll hook you up.

 

[Edwin-S]

I can get you Facebook. Cheap. I know a guy. LMK. I’ll hook you up.

Sorry, man, it.could never be cheap enough. Mark Zuckerberg would have to pay me. The only thing even less desirable is Elon Musk's Twiiter or, as I like to call it, shitter.

 

[BobO'Link]

Sorry, man, it.could never be cheap enough. Mark Zuckerberg would have to pay me. The only thing even less desirable is Elon Musk's Twiiter or, as I like to call it, shitter.

It was rebranded X for Xtinct.

 

[Clinton McClure]

I often wonder why hosting services don’t use RAID arrays and fault tolerant servers.