A this time - the site is SLOWWWWWW

[hammersley]

Wow this is worst than dial-up right now -----

 

[Steve St.Laurent]

It must have been something with either your ISP, your connection to the server, or your PC because the servers are smokin now. Have you scanned your PC for spybots? I recommend running both ad-aware and spybot s & d on a weekly basis. Here are the server loads for the hour previous to your post (in GMT time):



Web server:

3:30PM up 14 days, 5:21, 0 users, load averages: 0. 17, 0. 26, 0. 27

3:35PM up 14 days, 5:26, 0 users, load averages: 0. 70, 0. 39, 0. 31

3:40PM up 14 days, 5:31, 0 users, load averages: 0. 38, 0. 39, 0. 33

3:45PM up 14 days, 5:36, 0 users, load averages: 0. 47, 0. 42, 0. 35

3:50PM up 14 days, 5:41, 0 users, load averages: 0. 16, 0. 30, 0. 32

3:55PM up 14 days, 5:46, 0 users, load averages: 0. 41, 0. 35, 0. 33

4:00PM up 14 days, 5:51, 0 users, load averages: 0. 23, 0. 29, 0. 31

4:05PM up 14 days, 5:56, 0 users, load averages: 0. 25, 0. 25, 0. 27

4:10PM up 14 days, 6:01, 0 users, load averages: 0. 26, 0. 22, 0. 24

4:15PM up 14 days, 6:06, 0 users, load averages: 0. 75, 0. 43, 0. 32

4:20PM up 14 days, 6:11, 0 users, load averages: 0. 51, 0. 39, 0. 32

4:25PM up 14 days, 6:16, 0 users, load averages: 0. 20, 0. 32, 0. 31

4:30PM up 14 days, 6:21, 0 users, load averages: 0. 29, 0. 37, 0. 33

4:35PM up 14 days, 6:26, 0 users, load averages: 0. 46, 0. 44, 0. 37

4:40PM up 14 days, 6:31, 0 users, load averages: 0. 22, 0. 36, 0. 35



Database server:

3:30PM up 13:16, 0 users, load averages: 0. 48, 0. 33, 0. 24

3:35PM up 13:21, 0 users, load averages: 0. 10, 0. 18, 0. 20

3:40PM up 13:26, 0 users, load averages: 0. 24, 0. 16, 0. 17

3:45PM up 13:31, 0 users, load averages: 0. 12, 0. 17, 0. 17

3:50PM up 13:36, 0 users, load averages: 0. 02, 0. 08, 0. 12

3:55PM up 13:41, 0 users, load averages: 0. 33, 0. 15, 0. 13

4:00PM up 13:46, 0 users, load averages: 0. 20, 0. 14, 0. 11

4:05PM up 13:51, 0 users, load averages: 0. 81, 0. 57, 0. 32

4:10PM up 13:56, 0 users, load averages: 0. 90, 0. 77, 0. 48

4:15PM up 14:01, 0 users, load averages: 0. 55, 0. 80, 0. 59

4:20PM up 14:06, 0 users, load averages: 0. 10, 0. 34, 0. 43

4:25PM up 14:11, 0 users, load averages: 0. 19, 0. 20, 0. 33

4:30PM up 14:16, 0 users, load averages: 0. 20, 0. 25, 0. 32

4:35PM up 14:21, 0 users, load averages: 0. 22, 0. 25, 0. 30

4:40PM up 14:26, 0 users, load averages: 0. 15, 0. 19, 0. 24



as a comparison here's what the server loads were 2 days ago:

3:30PM up 12 days, 5:21, 0 users, load averages: 1. 51, 1. 19, 1. 02

3:35PM up 12 days, 5:26, 0 users, load averages: 2. 01, 1. 59, 1. 24

3:40PM up 12 days, 5:31, 0 users, load averages: 0. 63, 1. 04, 1. 09

3:45PM up 12 days, 5:36, 0 users, load averages: 0. 78, 0. 94, 1. 03

3:50PM up 12 days, 5:41, 0 users, load averages: 0. 63, 0. 68, 0. 87

3:55PM up 12 days, 5:46, 0 users, load averages: 0. 72, 0. 73, 0. 83

4:00PM up 12 days, 5:51, 0 users, load averages: 0. 93, 0. 85, 0. 84

4:05PM up 12 days, 5:56, 0 users, load averages: 0. 73, 0. 71, 0. 77

4:10PM up 12 days, 6:01, 0 users, load averages: 0. 57, 0. 64, 0. 72

4:15PM up 12 days, 6:06, 0 users, load averages: 0. 87, 0. 75, 0. 73

4:20PM up 12 days, 6:11, 0 users, load averages: 0. 91, 0. 86, 0. 77

4:25PM up 12 days, 6:16, 0 users, load averages: 0. 69, 0. 72, 0. 72

4:30PM up 12 days, 6:21, 0 users, load averages: 0. 80, 0. 75, 0. 73

4:35PM up 12 days, 6:26, 0 users, load averages: 0. 54, 0. 71, 0. 72

4:40PM up 12 days, 6:31, 0 users, load averages: 0. 53, 0. 68, 0. 71

4:45PM up 12 days, 6:36, 0 users, load averages: 0. 90, 0. 78, 0. 74

4:50PM up 12 days, 6:41, 0 users, load averages: 1. 28, 0. 95, 0. 81



2. 0 is a fully loaded server - meaning things are going into a queue. As you can see the servers are just loafing along compared to a couple days ago and these are the lowest loads I've ever seen on the server. The slow query rate for the last 10 hours (when I rebooted mysql) is . 007% or 7. 2 queries per hour taking more than 10 seconds to run. A couple days ago the query rate was . 043% or ~60 queries per hour taking more than 10 seconds.



If you see slow performance you may want to try a tracert by hitting start->run typing "command" and then typing "tracert www.turbodieselregister.com". That will trace your route all the way to our host facility and how long it's taking to get through each spot - if you see any asterisks it means that location is timing out and will slow things down tremendously. The last entry (69. 61. 23. 10) which is the TDR server will time out because I have pings blocked at the server - the furthest you'll get is global compass which is our host facility.

 

[hammersley]

Nope - everything is 100% here -

JUst had brand new install - nothing on my PC - and she smokes on other sites - but not the TDR - much better now than when I sent the message -

 

[JHardwick]

I'm here at 7:14pm, not quite prime time, but it seems faster to me without any changes to my system.



So far so good for me Steve!



Not that I'm a puter genious, or have ever maintained a server. But, doesn't it make more sense to use the horsepower of both servers all the time and only partition maybe a portion of each for backup? I dunno, just a question from a puter know-nothing. Sometimes applying common sense to a computer will get you in trouble, lol.



ie ... ..... clicking "start" to "shut down"

 

[msperos]

It's running great on the Quest DSL and the T1 at work great job Steve . Do you ever sleep

 

[Steve St.Laurent]

Try the tracert next time hammersley. If there is a problem on one of the connections between you and a specific website you can have trouble with one site and not on any others - because the traffic is going a different way. Right now for instance between my PC and the TDR server it's going through 10 different routers before it gets to the TDR server - a problem on any of those routers could slow it down.



JHardwick - think of the way I had it before as a standby server. It was setup such that if there was a problem with the primary server once I was aware that there was a problem I could have the site back up and running completely on the other server in about 10 seconds - it was just sitting there replicating the main server and could take over at any time. The way it's configured now if either server goes down I will have to reconfigure the other server to take over both functions which will probably take about an hour. Until the recent traffic explosion each server individually was plenty to handle the traffic of the site so the way I had them setup was the best option for using them and us having a maximum of up time on the system. In addition the way I had it when I was implementing major changes to the site I could just take the standby server offline and do the work on that and test it since it was an exact copy. I won't have that option any more either - any changes I'm making to the system will have to be tested on the servers while they are running the site as well. Initially when the servers were bought it was thought that we'd need them both. I took over webmaster duties shortly after those servers were purchased and once I got them configured properly it was obvious that we didn't need them both at that time. They wouldn't have even bought the 2nd server had they known that. It was nice to have the flexibility while it lasted though, I have had to switch over to the other server 3 times in the last 2 years. Just raises my stress level a bit now - I can deal with it .

 

[Yo Hoot]

Overall she seems faster.