MC/Service Guard

[Kevin Griesmar]

I am looking for information on MC/Serviceguard and possibly get some feedback from others who are currently using it at their company.

[melvyn burnard]

you could start by looking at the manuals at:
http://www.docs.hp.com/hpux/ha
There is also information available at http://www.hp.com

There are also books available, including one I believe is called Designing Highly Available Clusters

[James R. Ferguson]

Kevin:

If you want high-availability where one of your serves serves as a failover for the other, then MC/ServiceGuard software is what you want.

Various configurations are possible. An active/standby configuration is one in which a standby server is configured to take over after the failure of another. In this configuration the backup node may be idle or it may be running a less important function. An active/active configuration is one in which several nodes run critical apps while some serve as backups for others while still running their own applications.

Your hardware requirements depend on the particular configuration you wish to establish. At the risk of greatly oversimplfing, at a minimum your hardware requirements involve the addition of a "lock disk" and an RS232 "heartbeat" for the nodes to determine "aliveness" should the LAN fail.

An excellent starting point for understanding high-availability is an HP professional book "Clusters for High Availability: A Primer of HP-UX Solutions" by Peter S. Weygant (ISBN 0-13-494758-4).

If you are serious about MC/ServiceGuard, be sure to take HP's course on it before you start. Setting up a successful MC/ServiceGuard environment is not trivial and requires a good knowledge of hardware and software.

You can also read the manual at:

http://docs.hp.com/hpux/onlinedocs/B3936-90026/B3936-90026.html

This, along with the book I cited above, will certainly start you on your way.

...JRF...

[Kevin Griesmar]

I thank you both for your help.

[Rita C Workman]

OK you got the books and the info...you wanted to know about using it.

We have a 3 node cluster and will be going to a 5 node cluster. I highly recommend reading everything you can get your hands on. I like MC/SG, but it can be tricky (and frustrating) if you are not familiar with it. It is also something you want to carefully map out, cause when you have to make a change to one...it's a change to all. One of these days I may even get to a class on this. So far with the books and only one posting on this site about it, I have managed to keep it running.

So for an opinion on MC/SG...like I said..I like it.

[Stefan Farrelly]

Almost all of the sites I have worked at use SG. It is very popular and an excellent product. Most of the large companies who need minimal downtime use it. As long as its configured correctly (which can take some time and can be difficult for large multi-node clusters) and well documented it is simple and easy to use.