Fuck, I accidentally deleted a real post in the mess of spam that gets hammered. Sorry about that, I guess I’m a bit trigger-happy with the “DELETE ALL” button.

I came into work in the morning and, after ingesting a nutella-laden piece of potato bread, started ticking things off my to-do list. The data selections need to be ordered into columns, not rows, there should be breadcrumb links, options to order the final results by values instead of just location, etc. Unfortunately, the entire thing is instead spitting out

Mysql::Error: Got error 28 from storage engine: SHOW FIELDS FROM reports

Which was caused by something fairly obvious –

Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
                      224G  212G     0 100% /

Now, in FreeBSD this wouldn’t be a serious problem. I could just allocate a memory-backed filesystem with mdmfs, then union that over the filesystem that MySQL is attempting to write to (probably /tmp or /var). Unfortunately, Linux doesn’t include features like unionfs in it’s base distribution (at least, Fedora Core, the distro we’re using for our servers, doesn’t). And needless to say, I don’t want to start deleting random shit on a server I’m not overly familiar with.

The whole problem is aggravated by the fact that the traditional “use a separate partition for most of the base directories” was ignored when they set up this machine, so something filled up the entire disk and is making a general mess of things.

I almost prefer the rogue spinlocking Java processes that pop up every now and then. Honestly, it’s scary how poorly our department’s dev server is run, and we have no one to blame but ourselves.