MySQL Server Has Gone Away And Other Similar MySQL Nasties
 

Ok, what happened is this, my server is hosted with 1&1 and on Thursday of last week they descided to update their version of Python and the likes. My server and subsequently Photopost had been working smoothly and without errors up until this point, directly after this point, my Photopost is now spitting out 12 individual MySQL Error Report emails to me every time someone clicks on the "all" search button on the main PhotoPost front page, together with the search results page being full of errors. The rest of my PhotoPost appears to be currently working fine and as normal, just this "all" search button and only from the main frontpage of my gallery is reporting these errors.

I've already Google'd and searched on here for possible answers as to why this may have happened, but so far have come up with a blank, besides the fact it looks like 1&1 have made a major screw up and that they've so far taken 5 days of apparent "looking at the problems" without any kind of result at all.

Anyway, i'd like you to look at a small selection of the error reports and let me know if there is any possible way at all that this could be a series of Photopost script related errors brought on by my server now using Python 2.44, Python 2.5 Ruby Version 1.85 and Perl 5.8.8 and if not, please give me your best educated reasons as to what the cause of the problems are, as o far 1&1 are proving worse than useless and are seeming claiming to have fixed the server problem, even though I can not turn off maintainance mode on my website at all at the moment, due to this.

Please see the email error reports which give a combination of the following 3 errors:

Obviously there is more, but I think this gives a pretty good example of what is happening.

This isn't a request for help as such, unless it actually turned out to be something which is caused by PhotoPost which you could actually help me fix. I believe this is more a case of wanting to prove to myself that i'm not going mad here and that this is definately a server level problem caused by something being broken and not a script level problem as 1&1 at one point tried to claim.

If possible what i'd desperately like, is if someone could let me know which direction I should be pointing 1&1 with towards looking for the issue, as at the moment it looks to me like their support and techies don't know their ar*es from their elbows.

MYSQL HAS GONE AWAY means the mysql server is down. If it is not up you need to contact your host to resolve this issue as they have not properly upgraded their servers. You may occassionally get these errors if the host restarts the mysql server but this should not be a constant error.

Is there any possibility whatsoever that MySQL could go down or do some kind of weird time out under the pressure of a massive request? For example by clicking on the "all" button and it trying to return 100,000+ records?

Or would you know if it was for example possible for a memory setting to be set weirdly low, causing MySql to go away when a large request is entered?

I know the answers to both are pretty certain to be zero, but either this is some completely unique problem that is beating all the techs at 1&1, or they are being completely rubbish and not even looking for the problem at all, thinking it is this script causing the problem. Hence my serious need to narrow down the possibility completely to 1&1. 1&1 are now in to day 5 of supposedly looking at this issue you see and after 5 days of phoning and emailing them, i've had absolutely no joy out of them.

Your techs should be able to know what the error is by looking at their servers logs it would be impossible for me to suggest what is wrong with the server but they should know looking at error logs etc.