technical issue Equestria.tv 502 error

[Kivil 1,721]

Each and every day that I get on Equestria.tv ther is alway a 502 error. I get on my channel everyday because I co-own a Twitter, Instagram and soon to be Tumblr account and I post what songs/animations are playing. But that is off track. How do I fix this 502 error. It is stopping me from doing everything!!!!!

[Yozer247 1,006]

The 502 error is a server-side problem. It has been repeating for the past few weeks and the only way of temporarily fixing it is to restart the server every day.

They are currently looking into fixing the problem, so please bare with us as we try to figure out whats crashing eqtv.

I don't imagine it will be too long until the problem is found and fixed, however tech is also handling the migration of ponymc onto poniverse servers, so we are going to have some patience on this.

 

This will not effect any etv events such as movienight or episode streams. Although im sorry you guys can't access your channels when you want, ill check etv regularly and ping tech for a restart whenever I see it down.

[Kivil 1,721]

The 502 error is a server-side problem. It has been repeating for the past few weeks and the only way of temporarily fixing it is to restart the server every day.

They are currently looking into fixing the problem, so please bare with us as we try to figure out whats crashing eqtv.

I don't imagine it will be too long until the problem is found and fixed, however tech is also handling the migration of ponymc onto poniverse servers, so we are going to have some patience on this.

 

This will not effect any etv events such as movienight or episode streams. Although im sorry you guys can't access your channels when you want, ill check etv regularly and ping tech for a restart whenever I see it down.

It will be pretty hard for people to go to movie nights when we can't access equestria.tv. But thanks for the response.

[Jeric 46,835]

It will be pretty hard for people to go to movie nights when we can't access equestria.tv.

 

 

What Yozer was saying is that the time and date for movie night is static and a restart of the server can be planned long before the scheduled film thereby mitigating any outage. That isn't the case on other server outages. In those instances, when the system administrator is informed he will restart the server. 

 

So basically, no concerns with Movie Night or scheduled events. It would be the ad hoc usage that is impacted. 

[RainCode 61]

I know it's a little late to post because of the inactivity in this thread, but I do feel that I'd need to point something out.

 

I've noticed that the issue is on-going. In fact, I'm never able to access the site because of the gateway error.

Is the issue still being looked at? It had been almost 1 month since the second post was submitted (marked best answer) so it's kind of hard to tell on my end.

 

 

EDIT: It seems to be working now.

Edited August 22, 2015 by RainCode

[Yozer247 1,006]

I know it's a little late to post because of the inactivity in this thread, but I do feel that I'd need to point something out.

 

I've noticed that the issue is on-going. In fact, I'm never able to access the site because of the gateway error.

Is the issue still being looked at? It had been almost 1 month since the second post was submitted (marked best answer) so it's kind of hard to tell on my end.

 

 

EDIT: It seems to be working now.

 

Yeah the issue is still present. At the moment we have a setup that emails tech whenever it goes down, they then restart it as soon as they can.

However starting from this weekend the our tech guy feld0 thats been doing the majority of the resets has been in bronyCAN so today had an unattended 502. Lavo is now watching over eqtv so should reset it anytime it goes down.

 

The problem itself seems to be quite the tricky one, feld has setup some crashreport dumps to try to figure out whats going on to make the server flip out.

At this point all we can do is wait for tech to find out whats going wrong with the server.

 

Im sorry to hear you are experiencing major disruptions with eqtv, I thought the 502s have been taken care of because of the email system we have up, I guess our tech can't be awake to restart all the time

[RainCode 61]

Yeah the issue is still present. At the moment we have a setup that emails tech whenever it goes down, they then restart it as soon as they can.

However starting from this weekend the our tech guy feld0 thats been doing the majority of the resets has been in bronyCAN so today had an unattended 502. Lavo is now watching over eqtv so should reset it anytime it goes down.

 

The problem itself seems to be quite the tricky one, feld has setup some crashreport dumps to try to figure out whats going on to make the server flip out.

At this point all we can do is wait for tech to find out whats going wrong with the server.

 

Im sorry to hear you are experiencing major disruptions with eqtv, I thought the 502s have been taken care of because of the email system we have up, I guess our tech can't be awake to restart all the time

No worries at all. These sort of conventions can be so fantastic that one rather spend their time focusing on the event they've joined (especially if it isn't going to go back on for a while), I mean you could easily see how good it is from the photos and videos. 

 

 

Hopefully there's something useful in the log, they can sure be a mess! Quite interesting for it to be happening just on EqTV even though the resources allocated to it seem to be fine (according to the load speed).

 

So, you mentioned that there's a setup that would send out the emails if it were to go down.

I was thinking that if the emailing is done on the affected server then a reset script could be set up alongside the emailer so that when the 502's come along after a certain period of time (assuming that they can pop up for just a second) it could be done automatically. Of course, I don't know the set up and what's going down exactly so it's a bit of an assumption.

 

I'm looking at Google (face it, who doesn't) and there's quite a lot of results that could help (to do with php-fpm and process limits).

Is there anything set up to log or email the (active) processes along with the resource usage associated with each one? That could also help.

 

 

 

I know, I'm sounding like tech support now. 

 

 

 

Oh, last thing! When the 502's were around, the Poniverse status site reported it to be up. I had also noticed the past incident about it at http://status.poniverse.io/?start_date=2015-07-21 (18th). Did this have to be manually identified or is there an update timer in place?