Results 1 to 3 of 3
Thread: Linux HA clustering
Enjoy an ad free experience by logging in. Not a member yet? Register.
- Join Date
- Dec 2009
Linux HA clustering
I have a few high level questions with regards to this setup and i hope you gurus can enlightened me.
I have 2 nodes (A and B).
With Linux HA,
q1) Do i still need to sync data/applications between this 2 nodes or will it be sync automatically ?
e.g i got an app (myprog) running in background, if node A fail, node B will takeover by starting the app (myprog) on node B itself, but myprog must 1st exists in node B isn'it it ?
q2) I believe HA depends heavily on the heartbeat daemon, so therefore, if these related processes went down, won't my HA fail ?
where is these daemons installed ? on both node A and B ?
q3) I have read about certain apps would still need to be manually restarted or configured in a HA environment (warm standby)
I can't really picture out what kind of apps would require so in a HA environment ? (database?)
From what i see, if i have 2 sets of similar apps on both A and B, if A goes down, the apps will be start on B and that's it. (why would it still need any manual intervention ?)
Whats the software you are using to achieve HA?If you haven't decided on it checkout GlusterFS - AFR setup.
- Join Date
- Apr 2012
guide to get an idea of a practical example of Linux HA.
I didn't write it by the way, but I did help with some of the (poor) load testing.