User: Password:
|
|
Subscribe / Log in / New account

NLUUG: Minimizing downtime on servers using NanoBSD, ZFS, and jails

NLUUG: Minimizing downtime on servers using NanoBSD, ZFS, and jails

Posted May 13, 2010 16:55 UTC (Thu) by bronson (subscriber, #4806)
Parent article: NLUUG: Minimizing downtime on servers using NanoBSD, ZFS, and jails

Funny, I'm evaluating basically this setup using LXC and btrfs on 2.6.32. Verdict: not bad. There are some wrinkles, like LXC can't tell when guests have restarted or shut down, and sometimes it takes effort to figure out which guest is the one hammering system resources. I also expect that btrfs will get cranky once I go past 50% disk utilization but I haven't seen any problems there yet.

I'm I'm trying F12 (333MB), Debian Sid (211MB), and Ubuntu Lucid (152MB) as guests, all set up with debootstrap/febootstrap. They're all fairly bloated. I'd love something like NanoBSD but still feels like a regular Linux server.

Overall, it's been nice and stable. What to create a new guest? Just create a btrfs snapshot, copy a config file, and lxc-start it. Takes all of 10 seconds.

Not sure I'd recommend this for a production environment yet but I'm very happy with it so far. I'm excited to watch LXC continue to mature.


(Log in to post comments)


Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds