Hello,

the / -> Main_page redirect on en.o.o is broken again :-(

Additionally, suseROCKS told me that the wiki pages load slowly some days, however I can't reproduce that at the moment.

This is the first I have heard about wiki pages loading slowly in a very long time. In fact, it has been pretty snappy for me since the last upgrade. I'll keep an eye on it.

Full chat log: ...

And now to finally answer your previous mail:

Am Montag, 2. Januar 2012 schrieb Matthew Ehle:

...

...

Hello,

Am Dienstag, 13. Dezember 2011 schrieb Matthew Ehle:

...

...

...

>> Christian Boltz 12/13/2011 10:58 AM

Matthew, you probably already know that

http://en.opensuse.org/

(without specifying a page) is totally broken - I can wait forever, but nothing happens.

I was aware that it used to be broken, but I thought I had it working.

It seems this is still not (permanently) fixed - the same problem happened again yesterday evening[1]. (See also Rajko's mail.

I'm opening a bug report on this. I have never seen this before, and it only happens on that particular URL. It's very strange.

Can you CC me on it, please? (I'm suse-beta AT cboltz de in bugzilla.)

I sure will.

...

...

The situation is quite annoying - someone notices issues with the wiki (like this one), but you are our single contact (and SPOF). While you are doing a good job, you still need some sleep - which means it can take hours until you see a "HELP!!!!" mail.

I'd guess there's always an admin in the datacenter who could restart the reverse proxy or do other quick fixes, but we don't know how to reach him/her. Can you give us a mail address of the datacenter that is checked 24/7 so we can get wiki issues fixed even in the middle of the night [2]?

Please contact me directly. It requires a restart of the application, and a reboot of the servers is unnecessary and takes much longer. If you need me after hours, you should be able to call or text me on my mobile number. One of the SUSE employees or I can PM you the phone number. I am taking responsibility for this issue (plus I am in the on call rotation for web) so please don't hesitate to call.

That would be an option, but I'd say you should give your number to 10 or 20 people around the world.

Maybe sending your number to admin@opensuse.org would be a good first step - I'd guess the people there are trustworthy and there's always at least one of them near the keyboard ;-)

That sounds like a good idea. I'll do that.

Additionally, it would be nice to have a way we can use for free - I don't even want to know how much I'll be charged if I call or text you from germany... (Are there some websites that offer to send text messages to your phone for free?)

Maybe I can have the help desk forward this specific incident to me right away. You get a toll free number, and I get contacted as soon as someone notices it.

And BTW, even if a reboot takes longer than restarting the application, the total uptime of the wiki would probably be better with a reboot because of the faster reaction ;-)

Not really, because their policy is to contact application engineers before rebooting their servers. Also, they have an on call rotation just like the applications team does, so I don't see how it could be any faster.

...

I only ask that you don't give my number out :)

Not? Too bad - I hoped I'll get a number that I can enter at win game pages etc. :-(

Why not? Guess who they contact if someone ever does win at that? :-)

...

By the way, did I mention that it's fixed for now?

Unfortunately the fix doesn't seem to be permanent :-(

It is temporarily fixed again. I'm going to try a more involved approach by deleting and recreating the English wiki proxy service. The English wiki will be completely down for maybe 10 to 15 minutes.

What about a monitoring that measures how long it takes to load the en.o.o/ -> en.o.o/Main_page redirect? If it isn't finished within 10 seconds, a notification (or even an automatic restart of the application?) would be a good idea[tm] ;-)

We are trying out some new monitoring software, courtesy of our NetIQ business unit. This would be a good place to start. I'll ask about it next week. -Matt

Hello,

...

the / -> Main_page redirect on en.o.o is broken again :-(

same problem again :-(

Fixed again :-)

...

That sounds like a good idea. I'll do that.

Did you do that? (Note that I don't receive admin@o.o mails)

I honestly can't remember, as it would have been awhile ago. If someone on the admin list needs my number, they can PM me for it. I can also provide the help desk number, although it's probably not worth the free call.

...

We are trying out some new monitoring software, courtesy of our NetIQ business unit. This would be a good place to start. I'll ask about it next week.

Is this monitoring active already? I somehow doubt ;-)

It's not active yet, but we have someone who is working on it. We are still learning its capabilities and how to use it. Access Manager will be one of the first products to receive monitoring, and I'll make sure that this specific use case is included (until we figure out a permanent fix at least). Speaking of a permanent fix, chances are very good that this will not be an issue in 3.2. The proxy and caching is all based on Apache, which is a much more proven platform. I'm personally very excited about this for a number of reasons. We should be upgrading within the next couple of months.