...since having an actual downtime thread would be a bit inauspicious I'm just going to apologise for Annie's lugubriousness this morning here. I'm not exactly sure what happened; the problem resolution method consisted of "restart apache2.service and mysql.service until things return to normal," which is rarely evidence of a well-comprehended problem. dmesg mumbled something vague about apache being out of memory, which makes me think we may've just hit on some bug—obviously whatever happened was not the usual big-page-related-issue. I'll look into any documentation regarding recycling of httpd processes if this ever happens again. (And it probably will.)