Linux Node - OS reinstallation

[Nicolaus-sr]

Respected Customers,

One of our Virtuozzo Linux VPS nodes in the subnet 91.186.25.* is down at the moment. We are having to reinstall the operating system on it owing to core file system inconsistencies. This is being dealt with the highest priority. We will update this forum with further progress.

[Nicolaus-sr]

Hello Customers,

We have encountered a new problem with the drivers during the installation. This is being started once again.

[Dan]

Cheers for the update Nick

[Nicolaus-sr]

Respected Customers,

The VPS node is now back online.The OS issues have been resolved completely. As always, we thank you all for having been so patient during this while.

[chrisofcov]

No it isn't back online and no it isn't resolved completely.

Quote from online chat,
Our admin is working on it, as we still have some problem with file system. We are running FSCK on server.

A previous post says this could take between a few hours and a few days to complete.

Splendid.

[Nicolaus-sr]

Hello,

This was resolved. However, the hardware node has suffered the same problems once again. This is the second instance during the course of the day that this has occured. We are investigating this fully as informed on live chat.

[Stanley]

Hi,

I run a lightly loaded VPS web hosting on this node - it provides DNS & SMTP services + MySQL backend for virtual users. It runs pathetically slowly (and always has) - especially when compared to my other VPS on a different server that does a lot more !!!.

I would suggest looking a lot harder at the problems being experienced on this node.

And 5 hours to run fsck ? get real guys.

Twice in one day ?



Regards,


Stan

[Nicolaus-sr]

Hello Stanley,

While I agree with your sentiments, FSCK can sometimes take upto days. As I said, we are yet to know for certain what caused the node to go into FSCK for the second time.

We are also looking very closely at the hard drives and if need be, we will have them replaced.

[Stanley]

Hi,

And this node is still having problems :-

jasmine:~# time ps auxwww
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
root 1 0.0 0.0 1868 656 ? Ss Jul15 0:00 init [2]
daemon 31893 0.0 0.0 1604 352 ? Ss Jul15 0:00 /sbin/portmap
...root 14130 0.0 0.0 3784 944 pts/0 R+ 12:47 0:00 ps auxwww
dovecot 14155 0.0 0.0 3216 992 ? S 12:47 0:00 pop3-login

real 0m23.207s
user 0m0.001s
sys 0m0.008s

23 seconds to run a ps auxwww, my other VPS (different node, much heavier use) 0.66 seconds which is what I would expect.

jasmine:~# time uptime
12:49:34 up 17:55, 1 user, load average: 5.86, 2.04, 0.79

real 1m27.056s
user 0m0.000s
sys 0m0.002s

A load average of 5.86 and over a minute for uptime to complete !!

jasmine:~# time df -h
Filesystem Size Used Avail Use% Mounted on
/dev/vzfs 10G 2.1G 8.0G 21% /
tmpfs 3.9G 0 3.9G 0% /lib/init/rw
tmpfs 3.9G 4.0K 3.9G 1% /dev/shm

real 1m0.536s
user 0m0.000s
sys 0m0.005s


jasmine:~# time uptime
12:52:47 up 17:58, 1 user, load average: 2.69, 2.09, 1.03

real 0m2.437s
user 0m0.000s
sys 0m0.002s

Two seconds for uptime.

jasmine:~# time df -h
Filesystem Size Used Avail Use% Mounted on
/dev/vzfs 10G 2.1G 8.0G 21% /
tmpfs 3.9G 0 3.9G 0% /lib/init/rw
tmpfs 3.9G 4.0K 3.9G 1% /dev/shm

real 0m7.240s
user 0m0.001s
sys 0m0.001s

Notice the second run of df -h & uptime, both much quicker than the first run. This would seem to indicate that the binary was already resident so it was ready to run straight away, the first run that took over a minute would indicate that the binary had to be retrieved from disk.

Either the node is overloaded or there is still a problem with it.

top shows the machine at almost 99% idle.

Regards,


D

[karimali831]

Hi,

Please look at 92.48.108.38. I get a page load error

Thanks!

[karimali831]

Never mind no need to look up anymore, I contacted live chat and they resolved my problem straight away

[WH-Mike]

Hello,

We are again facing some file System on the server. We are working on the server and we will keep you updated.

We apologize for the inconvenience caused.

[chrisofcov]

Inconvenience?

Only slightly...

On Tuesday when the server broke you lost well over half my site and now you have done it again.

This time you managed to wipe out all database changes since 17-07 as well.

Keep up the marvellous work!

[Stanley]

Well said Chris - I was about to post and ask what was going on. You have my sympathies.

I did point out a few days ago that the host was still having problems but it looks as if my message was ignored. If they had bothered to look, they may have avoided this problem or at least been ready for it.

As it is, I now have a primary name server and mail server down and no idea of when it will come to life. My only bit of good fortune is that my web server is still up and running otherwise my phone would be off the hook :-(

Looking at the amount of time lost on this, they would have been better off building a new server and moving all the VS to that - oops there goes another airborne pig :-(

Regards,



D

[chrisofcov]

My mail server is up and running, delivering me mail from as far back as January?

After they ran the FSCK on Tuesday they knew then that the hardware was failing, but to quote support "there were no dirvers available for that hardware" so it could not be replaced. To me it's more likely time and cost which halted the needed fix.

I would like a definite answer whether they have bodged the server together as another temporary fix or have actually fixed it properly this time before I go and repair what I can on my site, I'm not wasting more time fixing things for them to go and loose it all again, and expect it all to be fine with a meaning less "sorry".