I have talk to our host again about this crash as it has happened a couple of times this past month and he has sent me this
As you can see, the load average went to 93 (when normally we'd expect it to be under 5). The servers memory became completely exhausted. This all happened within the space of 5 minutes or so.
From what i can see, the issue isn't actually related to a specific account or request, but the sheer number of simultaneous processes on your server - see the 'Tasks' count. 922 tasks (or processes), where at this exact moment on the server now it's under 200.
The question is, are these sudden 900 simultaneous requests legitimate traffic, or is it a problem with your site perhaps looping back and requesting itself for exmaple?
The last coupld of crashes have happened at 9pm and 4.30am which is hardly peak traffic for B2B websites. Is their anything that can be done - i have enclosed a screenshot of the top output just before crash