cowboy

At your SSH prompt check your version of CPAN by the command -- cpan -v
Version 1.81 had a bug that looped the perl installer. Also caused e-mail problems.

__________________
Where would you be if you were at the highest court in the land (US)?

Skipdawg

CPAN is version 1.87 cowboy

__________________

Skipdawg

Shoot it happened again! Server load this time shot over 9
Reopened the ticket on the server load issue so hopeful that got someone on it in time to see it and do something about it. This is getting really old very fast.

__________________

cowboy

I personally think this is just normal updates and backups which want to run full force and finish as soon as possible. They should last for less than a few minutes.

A load of 9 on a 4 CPU system is just over twice as long to do a job, which is still just milliseconds. I do not get concerned on my server with a double or even a 4x load, as long as it does not last extended periods of time.

The CPAN script I referred to last night raised loads to near 100 for hours.

__________________
Where would you be if you were at the highest court in the land (US)?

Skipdawg

Well support sure can't find anything that is for sure so giving up on that route I guess I'll just have to deal with it till I don't feel like paying for the annoyance anymore.

__________________

state

Hey Skip,

I'm having the same problem you are now. I was moved to a new node when I upgraded my VPS package and have had nothing but troubles. My resources keep showing exhausted when nothing is going on, last night my VPS rebooted several times by itself.

The VZ panel even showed that I exhausted my inodes this morning around 1am which seems pretty impossible since I wasn't even doing anything with the site at the time (and I certainly don't have a cronjob running creating files and sym/hard links in the background!)

Constantly the load has been so high that webpages, mail, etc don't even work. It took me over a minute and a half to execute a simple 'w' command from the shell with this depressing result:

Paul has been looking at this issue for me, hopefully he'll isolate it soon. That ticket number for those interested from support is UUQ-572337.

I wonder if we can get credit back on service (as with an ISP) for outages?

Skipdawg

If Paul finds something let me know. I gave up on them telling me was all fine. One can only listen to that so many times when they know it's not fine.

It has simmered down a little for me but Monday night I had some heavy something going on and nothing was showing as being used?

But I really like my VPS. If they could solve and fix this issue I'd probably love it shoot maybe marry it LaughLaughLaugh

__________________

state

Will do. For the amount of money that I'm paying per month I've no plans to drop this anytime soon.

I started poking around trying to figure out what other hosts were on the same VPS as I am (using 'netstat -an' and the whois.sc reverse-lookup tool) and found several. They loaded OK when my site (and the load) were OK.

Just moments ago the load shot up again and my sites took several minutes to load. I used that opportunity to check the reachability of the other sites and they were timing out as well. Something is terribly wrong in VPS land.....

Skipdawg

Well this is around my norm 22:29:12 up 106 days, 1:40, 0 users, load average: 0.11, 0.07, 0.07 and very low CPU usage, most of the time below 10%

So when the server load shoots to over 9+ and the CPU spikes out at 100% and it says nothing is doing it. Yea sumtin ain't right in VPS land.

I noticed when I do catch YUM or Mailman active the server load is like between 6 to 8 which is not to bad if less than 5 minutes.

But when nothing is showing as using the resources and it jamms away for abit and support dose not see it grrrrr The Dawg wants to cut loose. hehehe

__________________