hh, too many forks, PIDs got reused, we’re confused

I’m not entirely sure I’m in love with EXT4. I’ve installed OpenSuSE 11.2 on my laptop (64bit, SMP) and I said “WOW EXT4 IS OUT LETS USE BLEEDING EDGE FILESYSTEMS WHAT COULD POSSIBLE GO WRONG?”

Everything.

This laptop has a slow, slow harddrive, so I figured any filesystem doing bleeding edge fast stuff might help. My previous favorite was XFS which I’ve had nothing but fantastic luck with. EXT3 was also nice on the servers and was bulletproof, although noticeably slower than XFS. EXT4 is fine until you put the filesystem under load. It’s blazing fast until it runs out of cache, and then you’re screwed hard waiting for it to catch up. In my case I managed to blow it by coping my documents directory I backed up to my other PC back to my laptop (about 6GB worth of crap) while running X, firefox and setting up YAST update sources. I thought things were going way too well when we hit the commit wall. The drive was on solid, and I could move the mouse. Clicking on things resulted in the drive chugging for a bit and then slowly starting the animation. Nothing ever launched. Eventually I couldn’t even kill X. I ended up pulling the plug.

The good news is that it recovers gracefully from a crash. The dreaded “empty file” error didn’t happen to me, but I’m marginally pissed the filesystem will let you outrun it catastrophically. That being said, when I restarted the SCP, I got the cryptic error: hh, too many forks, PIDs got reused, we’re confused…

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s