If you can't get a shared primary IP then DNS failover is pretty much
your only option. While you can do shared primary IP across two
locations like this it usually involves a third leg, so the DNS
solution is probably simpler.
To fully automate you would still need some kind of high-availability/
failover script running on the slave to detect when the primary server
has blown up. It would then activate the standby server and tell your
DNS servers about the change (at which point you're at the mercy of
DNS propagation & caching, so it's not an "instant" switch, but if the
TTL on the intermapper record is short it should be pretty fast).
This is pretty easy to rig up on *NIX systems, and I vaguely recall a
Microsoft-Approved tool that can do the same thing (If I'm wrong about
the MS tool a little VB or PowerShell scripting should be able to do
the job too).
If you're not already running it you may want to look into InterMapper
Database as well: you can set up replication between two Postgres DB
servers ("decent" replication exists right now, "good" replication
coming soon in 9.0) which will preserve all your data (charts, etc.)
so you don't have gaps later if you do routine reporting.
I know maps & devices are defined in the IM DB schema too, but I don't
recall if you can easily pull them back out -- If you can this pretty
much automates the file-copying part of your routine & only adds a
little more complexity to the fail-over (You would need to activate
the standby database & pull the maps out of the DB prior to starting
InterMapper).
If the map-extracting functionality doesn't exist yet it might make an
interesting RFE from a disaster-recovery standpoint.
-MG
On Feb 17, 2010, at 6:59 PM, Kevin Wigle wrote:
thanks for the suggestion but we are a Microsoft shop.
Also the backup is in a building across town so the "shared primary"
wouldn't be on the same subnet.
thanks again
Kevin
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of
[email protected]
Sent: 17 February, 2010 5:36 PM
To: InterMapper Discussion
Subject: Re: [IM-Talk] Backup - Redirection
Kevin, perhaps configuring your OS with hot failover scripts would
work well for you. Most *nix flavors can be pretty readily configured
to monitor each other and take over answering for a 'shared primary'
IP.
Randy Millsop
-----
Sent from my iPhone
On Feb 17, 2010, at 1:27 PM, "Kevin Wigle" <[email protected]>
wrote:
It is that time of year again that someone asks the question about
disaster recovery and Intermapper.
I constantly backup the Intermapper Settings folder but "they" want
a more involved solution.
We have a backup Intermapper that I update constantly but it is
turned off waiting for the main to crash.
So, "they" have just figured out that none of this is automatic.
My question of the day is how to re-direct connections to
Intermapper at x.y.100.50 to x.y.101.50
That would be any and all requests (not just http) that arrives to
the crashed server should be sent to the new server but without the
users having to do anything.....
I've seen hardware and software things to do this.
Any suggestion on something simple?
Kevin
____________________________________________________________________
List archives:
http://www.mail-archive.com/intermapper-talk%40list.dartware.com/
To unsubscribe: send email to: [email protected]
____________________________________________________________________
List archives:
http://www.mail-archive.com/intermapper-talk%40list.dartware.com/
To unsubscribe: send email to: [email protected]
____________________________________________________________________
List archives:
http://www.mail-archive.com/intermapper-talk%40list.dartware.com/
To unsubscribe: send email to: [email protected]
____________________________________________________________________
List archives:
http://www.mail-archive.com/intermapper-talk%40list.dartware.com/
To unsubscribe: send email to: [email protected]
