De Leek!
I have a resource leak on one of our live servers. It’s not seriously serious, but it’s certainly seriously annoying. The box in question is running a dozen or so pretty busy services and has an uptime of a shade over 577 days. And something has it 270K into the swap. Not a big issue, no. There’s plenty of physical unallocated buffer RAM available. There are no overly large processes. There’s just 270K. Sitting there. Lurking. In the swap partition. Except tomorrow, it’ll likely be 270K. Cos it grows. And lurks.
A reboot would fix it, but I aint rebooting a box with that kind of uptime for a 270+(n, n=number of days after today)K annoyance. It’ll be fecking perl, I knows it. All the cronjobs are terminating properly. There are no stuck or zombie processes. Everything is behaving as it should. If I turn off swap and then bring it back up again. It goes away. But then it comes back. Starting at 270K. Building by 1K per day. Every day.
And I swear it’s smirking at me 🙁
Leave a Comment