CCGI errors

[#1]

I am running a script on CCGI server which has been operating for a long time, and just started getting errors:
Warning: session_start(): open(/tmp/sess_9007155a88e0f6cc8375890dab61dab4, O_RDWR) failed: No space left on device (28)
I am thinking this is an issue with the /tmp directory
Would anyone care to comment please?
Thanks
Paul

[#2]

My only comment is that it is happening to me as well.

[#3]

Looks like one of the network drives on the CCGI server has run out of space. You both need to raise a ticket and include the error message.

[#4]

I've raised a ticket [23530785]
Still happening this morning!
Paul

[#5]

Is there a limit on the size of the database you can have in SQL?
I have used 150MB of my cgi webspace and have started getting errors using my eve-o killboard that refer to not being able to save the datbase.
Warning: session_start(): open(/tmp/sess_dca982cbe3df60d1b0950a9848b82c23, O_RDWR) failed: No space left on device (28) in /share/storage/02/sc/"username"/shado/common/class.session.php on line 26
Thanks for any help
Merrick

[#6]

Moderators Note
Threads about no space left for temporary sessions merged.
Jonathan
Also two posts by Jelv removed as they just pointed to the other thread.

[#7]

Quote from: chillypenguin

Moderators Note Also two posts by Jelv removed as they just pointed to the other thread.

Fine except you've also removed my hint that the second poster should raise a ticket! 

[#8]

ah cool, well at least someones on it

[#9]

It appears at present nobody from Plusnet is on to it. That's why you need to raise a ticket.

[#10]

The responses from CS were:
1.  I will forward this to our network engineers.
(INTERNAL)
please advise, appears to be a server problem
2. The following comment was added to the Question
(INTERNAL)
monitoring systems suggest diskspace is ok..
This may be the case - but what is causing the problems then ?
The common 'No space left on device (28)'  must point to something.

Paul

[#11]

Reply received from CS 4th December 2007
'I have found the cause of the problem, our monitoring did pick it up, however it was deemed to be a transient problem and so was not raised to our engineers. The root cause was that the /tmp volume (which is on the shared storage) ran out of inodes a number of times from friday onwards.
I have increased the number of inodes available on the volume by 25%, we will continue to monitor this and will increase it again if necessary.'

Thanks
Paul

[#12]

ah a solution! excellent. my problems stopped a little while ago but still wondered why ti was happening.