Holy system load Batman!
bc. 8:53AM up 67 days, 18:50, 2 users, load averages: 277.89, 233.78, 199.75
Holy crap! That was the load when I logged into chaos.amber.org this morning. It, according to some further records, skyrocketed close to 300, but I could no longer do anything through my SSH connection. For whatever reason, it then fell back down to normal about 10-15 minutes later. As far as I can tell, and from discussing with James, it appears to be Sawmill going insane. This is troubling because it has tended to be a very reliable product in the past, and blows everyone else on the market away in feature-performance-functionality. We’ll see if I stay with it, though. Maybe time to write something of my own.
It’s a testament for FreeBSD that the system took this level of load, and kept functioning long enough for the process to quit and everything to return to normal. Now if I could just get some real resource shaping in place so I can always have CPU for SSH. Alas uptime isn’t that impressive as the site it’s located at has had quite a few long power failures which drain the UPS.
This entry was posted at 9:35 am on 1 February 2005 and is filed under Meta. You can follow any responses to this entry through the post-specific RSS 2.0 feed.
No comments found.
Responses are currently closed, but you can trackback from your own site.