Have just now been able to sign on here for the first time since Sunday. Anyone else have a problem?

(12-11-2017 09:41 PM)usmbacker Wrote:  Have just now been able to sign on here for the first time since Sunday. Anyone else have a problem?

I could not all last evening and then this morning was able to.

(12-11-2017 09:43 PM)UCGrad1992 Wrote:  

(12-11-2017 09:41 PM)usmbacker Wrote:  Have just now been able to sign on here for the first time since Sunday. Anyone else have a problem?

I could not all last evening and then this morning was able to.

I just read the post above about the issue. I am AT&T so that was my issue it seems.

If I had to get somebody has been tinkering with dns routing statements on the server and some providers have slow refresh rates. Just a guess

(12-11-2017 09:49 PM)usmbacker Wrote:  

(12-11-2017 09:43 PM)UCGrad1992 Wrote:  

(12-11-2017 09:41 PM)usmbacker Wrote:  Have just now been able to sign on here for the first time since Sunday. Anyone else have a problem?

I could not all last evening and then this morning was able to.

I just read the post above about the issue. I am AT&T so that was my issue it seems.

That was my issue, too. Thankfully, it's all good in the 'hood, G. (Or something like that. lol)

The internet provider out here sucks so my coverage is spotty at times too. Couldn't some of the issues be weather related?

I am on Suddenlink, and could not get on either.

I tripped over the cord. Sorry.

I am still not being able to sign on at times.

I was AT&T and it was spotty at best. It would come and go last night.

I know GTS had to reboot the server Sunday night for maintenance, so that might have been it. It blew up the search function here, but I believe that's fixed now too.

I still couldn't sign in last night until I switched from 4G (Verizon) to wifi.

(12-12-2017 09:22 AM)UTSAMarineVet09 Wrote:  I was AT&T and it was spotty at best. It would come and go last night.

Same for me this morning,

comcast has the same problem....

also, gts apparently does a reboot at 4 am AND 4:30 am that usually takes about 10 mins. to clear....

I don't have fingernails anymore, got a lot of protein though. I thought the FBI or Russians had gotten ahold of this site for subversive, ahem, items.

(12-12-2017 10:01 AM)stinkfist Wrote:  comcast has the same problem....

also, gts apparently does a reboot at 4 am AND 4:30 am that usually takes about 10 mins. to clear....

LOL. Not quite, but your eyes do not deceive you. The search feature on this site depends upon some aggressive search caching. It's *VERY* computationally expensive to update, so I only do it once a day at 4 AM ET. There's just too many damn posts to search just using a regular search on the database. If you search for something with no/few results, just that search could take down the entire database server as it has to lock the post table to search all however-many-million-rows it is now. So it's not doing a "real" search, it's searching against that cache which is updated once a day. I'm locked into that caching system because this board does not support the very very VERY new features added to the most recent versions of MySQL to do this sort of thing without the aggressive caching. So this is yet another one of the big reasons why I took the rather severe step of writing my own replacement software for the board (Aurora). There's a lot of technical sausage making that isn't pretty to make the site run like this on off the shelf software. Looking forward to ditching that and launching Aurora next year.

Anyway, the super slow response you get then at first is the cache updating new posts. The SECOND super slow response you get a few minutes later is it merging those new posts into the master cache system. Both operations are ENORMOUSLY computationally expensive and essentially max out the CPU on the database server.

(12-12-2017 09:51 AM)Redwingtom Wrote:  I know GTS had to reboot the server Sunday night for maintenance, so that might have been it. It blew up the search function here, but I believe that's fixed now too.

This was a legit reboot from me. The server uptime was like 270 days or something huge like that, and my monitoring software was starting to raise warning flags about some flakiness going on with the Linux kernel itself (blah blah NMI watchdog alerts, some CPU cores showing skew, etc). Probably a very slow light memory leak bug in something, but easily fixed with a reboot once a year or so. And when it came back up from that reboot, the connection to the search cache on the database server didn't come up properly, hence you getting that "connection refused" error on search you PM'ed me about (thanks again for that).