Aaarggghhh.
After 20+ back-and-forth messages with WebIntellects, my hosting provider, the problems seem to have subsided.
They finally set the permissions so I could access my database, and I was able to restore the lost month of data from my backups, and get things back to where they were yesterday afternoon.
Except…I suddenly ran out of disk space. Which didn’t make any sense, because the only thing on this server is the blog, which isn’t that big. I increased the disk allocation from 100MB to 300MB in my reseller panel (I manage multiple domains from one account), noting to my surprise that I was using 175MB of that space, and tried again to update the database…only to get another space-related error. A check of the control panel showed I was now using all 300MB! Clearly a process had run amok. But I have no access to processes, so I couldn’t list them, let alone kill the responsible party.
After a few messages back and forth with tech support (through an annoying trouble-ticket system), I determined that the file that was growing so quickly was the error log. When I peeked at it, I found that it was the blacklist.pm module from MT-Blacklist that was cycling, adding hundreds of lines per second to the log.
I deleted all the MT-Blacklist files, and then had the tech guy kill the process and delete the log. Once I was sure comments worked again, I went in and tried to reinstall MT-Blacklist, but I got errors about undefined arrays. I’ve got a query in to the host about whether they’ve changed somethign about the perl install on the new server. In the meantime, I’m keeping comments closed on posts more than 30 days old, and hoping not to get hit too badly with spam between now and when I can get things running again.
www.flickr.com
|
