What The Hell...

[RMSSF]

Can someone please explain me in details what the hell happened to my VPS and all data inside it???

The IP is (was) 87.117.246.33

Thank You!

Rui

[Administrator]

Hello Rui,

One of our VPS Team member is looking into this and I will get back to you as soon as I hear from them.

Please allow them some time to figure out whats wrong with your VPS.

[Administrator]

Hello Rui,

You were contantly using 96% CPU resources of a high specification 2 x Quad Core Xeon Server. Applications you are running on the VPS wont run even on a high specification Dedicated Server we have.

you will need to move your VPS somewhere else.

[RMSSF]

I have been checking from time to time the VPS if everything was ok.
I have never noticed any high cpu load and I even believe that it has been using one of the lowest cpu, mem and traffic usage profile!
It has always been working just as a simple web server in proxy mode to a backend server.
Maybe then sudenly for some reason some application crashed, and if that was the case, whats next? You shut it down simple and quiet? Without even a warning?
I suppose the tech saw what application was causing the load since he changed the root password before shuting it down. Can you ask him what was causing the load? And why he had to shrink the filesystem space from 10Gb to 2 GB?
Honestly I believe that what happened was a mistake from one of your techs.
Where is all the data? At least I want the old data to move to another VPS!

Thank You.

Rui

[Administrator]

Hello Rui,

Please open a ticket with our VPS Support department from http://support.webhosting.uk.com and ask them to provide a backup of your VPS. Earlier your VPS was hosted on one of our OpenVZ nodes and we were not able to track your usage on that node. Now we have migrated most of our OpenVZ nodes to Virtuozzo and thats why we can track such usage efficiently.

You have used over 90% of CPU resources constantly as the graph shows 24 hours usage. Your root password was changed as we were trying to figure out what applications were overloading. All your data is safe and you will get that from the backup.

[RMSSF]

Again I insist there must have been some mistake.

My VPS has been always on a Virtuozzo node from over a year!
I know because I have always been logging in to VZPP from time to time to check satus and QoS alerts.
I've checked several times this month and everything was ok!
It had also installed a monitor tool configured to send me an e-mail if something is wrong, like when a critical service is stopped or the load is high.
Honestly it is not easy for me to believe that it has been using all cpu for so much time like you say.
One more thing, I had configured syslog deamon to send syslog messages in real time to my local system, and the only relevant messages I've received during this month was when the system was recently restarted and when all the services were shut down.
I could take notice 2 days ago when the system was still alive with all services running and looking in good health (not under high load) but I couldnt login as root already, and I also noticed the VZPP was not working anymore so I guessed it was moved to another node in the recent days, I have also noticed the filesystem space was shrinked to just 2GB.
Then yesterday VZPP started to work again but I still couldnt login as root.

Honestly all this situation does not look like if my VPS was having any kind of problem!
And I still cant understand how it is possible all this happening without me receiving a single notice from the staff about any problem with my VPS!
I cant understand why if in the case there is some problem with some application causing high load, you just shut down the whole system, kill and erase the VPS and your customer, all without sending any warnings or trying to fix the situation with the customer!
And I also cant understand why my package is marked as still active in the billing system!

Very interesting situation here.

Thank You!

Rui

[RMSSF]

Originally Posted by Administrator

Earlier your VPS was hosted on one of our OpenVZ nodes and we were not able to track your usage on that node. Now we have migrated most of our OpenVZ nodes to Virtuozzo and thats why we can track such usage efficiently.

Maybe the node moves was what was causing the system to crash??
My VPS was installed and configued on a Virtuozzo node, and it had special configuration to work with the TUN device (/dev/net/tun).

I cant see any other reason or situation that could make my VPS to crash.

Rui

[RMSSF]

I have some more words to say for now.

If I would have noticed that something was causing high load in the VPS, I would have taken immediate action to find and kill the application causing high load, or shutdown the VPS myself, without any problem.

As I said, it is not easy for me now to believe that my VPS was indeed with any problem of high load.
As the situation is right now, I cant check it myself if there is any problem, as I cant access the VPS.

So, I suggest, maybe you will let me check myself if there is any problem with my VPS?
If you agree, then you can let me know the new root password to access VZPP, then, at the appropriate time, I will login to VZPP and switch on the VPS and see if there is any problem. If indeed there is any problem and I cant fix it immediately, I will shut it down immediately again and contact support about the issue.
If you dont accept my suggestion, you will have to understand my disbelief and unhappiness about all this situation and about your methods to address the problems with your customers.

Thank You.

Rui

[Administrator]

Hello Rui, I have requested our VPS Team to unsuspend your VPS for next 24 hours. We will monitor performance of your VPS and node and we will let you know status of load on the node by tomorrow evening.

[RMSSF]

So you mean that I will get back the access to my VPS to check it myself?
If so then you can send me the new root password to my email, and maybe its better to leave it turned off until I login to check it.
And in any case I will turn it on only at late night so any problem with high load will have the least impact.

Thank You.

Rui

[RMSSF]

err... I see it is online already, but I still dont have the root password?
How can I see if it is under high load and check the problem if I cant have access to it? Even VZPP suddenly dissapears right before it came online?

Rui

[Dan]

Originally Posted by RMSSF

err... I see it is online already, but I still dont have the root password?
How can I see if it is under high load and check the problem if I cant have access to it? Even VZPP suddenly dissapears right before it came online?

Rui

The root password should be the same, if you are having trouble logging in contact Billing and get your password reset

[RMSSF]

Originally Posted by Dan

The root password should be the same, if you are having trouble logging in contact Billing and get your password reset

Hi,

Yes of course I am having trouble logging in, the root password was changed as I said before, and it was not still given to me, so I am having a little difficulty to check how things are with the VPS, and to make things even worse, VZPP panel is not available now, I cant figure why if my VPS was always on a Virtuozzo node and I always had access to VZPP.

Thank You.

Rui

[RMSSF]

I have contacted billing and they promptly gave me root again and enabled VZPP, so thanks to them.

So I have checked the VPS and saw what was going on...

They have screwed my VPS with the migrations!

There is not any faulty application, the system itself is screwed.

I cant even login through ssh:

Last login: Sun Jan 13 19:33:54 2008 from
-bash: error while loading shared libraries: libdl.so.2: cannot open shared object file: No such file or directory

This way they coudnt find what the problem was, even in VZPP it is reported that the load average values are "0.0; 0.0; 0.0" but it not correct because even that is not working anymore and you cant even read the system cpu load.

The problem is at kernel and node configuration levels, and my VPS configuration is screwed up.
Now I dont have any problem in believing that indeed my VPS was crashing the node, and so I have turned the VPS off and will leave it like that.

As it concerns to me, you can forget and throw away this VPS, and dont attempt to start it again unless you can fix the configuration at the node level.

Needless to say more, what it concerns to me, this issue is closed.
Pointless to explain about my indignation on how this situation happened and on how your staff have dealt with it from the beginning.

Worth to mention that I have never had any problems or issues with this hosting company before, and in fact I was enjoying and was very confident as their customer from the beginning after making the deal with one member of the staff at the time of subscribing the package, so I leave with sadness.


Thank You.

Rui

[Administrator]

Hello Rui, Please open a ticket for our VPS Support Team and ask them to reload your vps and restore your data once again. Not sure why libraries on your VPS have gone bad but we will get it sorted ASAP.