AletiaStatus.com

**tribby** · 05-27-2002, 01:24 PM

Very cool. Thanks Aletia!

**LeeUmm** · 05-27-2002, 02:41 PM

Sweet stuff.
I like it

**lookout** · 05-27-2002, 04:24 PM

Me too! Thanks also for the updates, Dana.

**eXodus** · 05-27-2002, 06:12 PM

Spent time on this, not customer support =/

**glo** · 05-27-2002, 06:29 PM

It's a start exodus, don't knock it.

**clio** · 05-27-2002, 08:23 PM

Dana's message said they were working on support too.
However, some questions:

- What do those green dots mean? What about the red dot? There's no legend on the status page. Pop on cobra is a red dot. Does that mean that that is not functional?
- You should link the systems to their actual system info page, which gives us more details

i.e. http://swan.nocdns.com/

- Why does swan have such high load averages?

**eXodus** · 05-27-2002, 10:12 PM

send in a support and i think aletia thought people would have comon sence red = bad green = good if there is a yellow = warning

**Julian Muñoz** · 05-27-2002, 11:24 PM

Uhmm I like it... Well. right now I see everything green... 6.7 load averages on swan... Open a support ticket swan users.

Well, I really like Aletiastatus.com. It gives the users a quick review of all the aletia's server

Ron Kuris · 05-28-2002, 09:41 AM

Originally posted by eXodus
Spent time on this, not customer support =/

If he's not getting what he needs, he should knock it. What ticket(s) are you talking about Exodus so we can look into them?

**lookout** · 05-28-2002, 06:42 PM

Actually, I thought the time well spent. Maybe not if there was some huge fire raging somewhere though, which I get the sense is Exodus's point. But even then, throwing people at a problem doesn't always resolve it more quickly. Having some people moving forward on other issues (like this) may actually have been more constructive.

I think the new report page a very handy tool for quickly verifying the status of any Aletia server. At least when some service like http or mail goes down, you'll have a quick way of checking whether the Aletia server is the culprit. Good info to know if you're about to report an outage.

It also should help tech support respond more quickly and effectively, which is generally good for customer service....

**eXodus** · 05-28-2002, 08:48 PM

I do nto have any tickets open currently, but when I do it always takes a while or i never getwhat i needed. Thats what i was refelring to.

**EHG** · 05-30-2002, 08:38 PM

Perhaps Ron is aware of this - then again perhaps not?

I just opened the pg:

http://www.aletiastatus.com

and being a duck with a curious nature I then opened the pg:

http://eagle.nocdns.com

The two pages are reporting a huge discrepency in reference to the load averages.

I took a snap of that part of each page within 5 minutes of each other the results of which may be viewed at:

snap of eagle comparing status pg w/ nocdns pg

Can't help but be a bit curious as to which is correct. Have doubts as to the reliability of the nocdns statistics anyway as eagle is still reporting a trillion mb's of memory or some outrageous figure.

JUST WENT BACK PRIOR TO POSTING THIS AND DID ANOTHER LOOK SEE AND THE FIGURES ARE ONLY SLIGHTLY DIFFERENT NOW - THEREFORE THE DELAY IN DOING THE SNAPS MAY HAVE CAUSED THE WIDER DISCREPENCY - I think I shall post this anyway perhaps someone will explain the load avgs to me

edit
Never say die! I opened both pages and sized them and then refreshed both almost simultaneously repeatedly. The two pages report from identical figures to a 100% discrepencie. Eagle has within 3 or 4 minutes had a load range from:

high of 5.00 2.31 1.27 to low of .39 1.35 1.07

and while doing this 1 user was being reported -

end of edit

I hate to admit ignorance however assuming and that is a huge assumption that we were able to get statistically correct data on the load averages of our server - exactly what does load average represent and at what level should one be concerned?

I also noted that users=0 on the nocdns page - Does users refer to those who might be accessing the server directly on aletia's end - that figure does not change when I either access my site or actually using ftp to upload or download files??

**tribby** · 05-30-2002, 10:06 PM

Originally posted by EHG
The two pages are reporting a huge discrepency in reference to the load averages.

I can take a stab at why that is. I'm nearly certain that aletiastatus.com is not updated instantly, but rather is a recurring process that gets updated every, say, 10 or 15 minutes.

So by visiting machine.nocdns.com you are seeing an instant grab of the 'uptime' command, whereas aletiastatus.com is a delayed view of the 'uptime' command. Thus, machine.nocdns.com will always be a little more accurate than the overview on aletiastatus.com.

Hope that cleared things up for you.

J

**EHG** · 05-30-2002, 10:28 PM

Originally posted by tribby

I can take a stab at why that is. I'm nearly certain that aletiastatus.com is not updated instantly, but rather is a recurring process that gets updated every, say, 10 or 15 minutes.

Don't think so!! If you go to the aletia status page and continuously refresh the figures change as they do on the nocdns page when refreshed.

I was somewhat surprised to see that the load averages could change so radically from literally one second to another (or as long as it takes for the page to refresh) so perhaps just that second or two between clicking refresh on one and then on the other is responsible and of course depending on the source info for each page might also cause a few second discrepency in gathering the info - BUT IT LOOKS SUSPICIOUS TO ME!

Also I agree with clio I think the nocdns page for each server should be linked from the status page which would make it quite a bit easier if one wanted furher info on their server.

Regardless still want to thank Aletia for this and anything else they are attempting to improve the quality of their service.

**jkeller** · 05-31-2002, 03:49 AM

Originally posted by EHG
[...] exactly what does load average represent and at what level should one be concerned?

The numbers represent the average number of waiting processes for the last minute, five minutes, fifteen minutes. Frankly they mean nothing on their own, but typically should be as low as possible. "0" is very nice, for example (meaning everything gets executed immeditaely, not that there's nothing running). A rule of thumb I've seen here is that you shouldn't exceed the number of processors times 1.5.

I also noted that users=0 on the nocdns page - Does users refer to those who might be accessing the server directly on aletia's end - that figure does not change when I either access my site or actually using ftp to upload or download files??

"Users" referes to the number of people logged in through ssh/telnet using the shell.