[kwlug-disc] Virtualization allocation?

[John Van Ostrand]

I think virtualization is useful. It saves on hardware costs and electrical 
costs, at least for those of us who use newer systems.

Virtualization partly exists to deal with server sprawl. Everytime many orgs 
needed a new service they obediently bought a new server and new OS license 
and created a new problem. Virtualization was the solution to the high cost 
of rack space, maintenance warranties, electrical and cooling costs and the 
cost of migrating the server to new hardware every so often.

The next problem around the corner, in my mind, is vm sprawl. Now that it's 
easy to create a new vm we have even more. Maintenance of these will be the 
next issue.

To summarize: splitting up simple services into seperate vms is still a 
problem. Having many more systems to keep updated and secure I think 
decreases security.

Not only that but you need precious ram and disk space for each instance. 
Instead of using say 2GiB of ram for one system that does it all you need 5 
x 2GiB. On big-ass machines disk and ram can get pricey.

We use virtualization because we want to use less resources. It also gives 
us the ability to more easily migrate vms to different hardware. Failure or 
just age may be the reason for the move.

Another value we see is that we are now running on diagnoseable hardware 
(ECC, etc.) So when a system fails we are more confident if it's bad 
hardware.

You may also consider the advantage of less power consumption in both server 
and cooling. You may consider the remote access feature: starting and 
stopping vms remotely.

For hardware for large vms external storage is a must, usually iSCSI or 
fibrechannel. Having it shared means being able to do things like failover 
to another host. Having it external (ie in a storage system) means security 
and on-line modifications and expansion.

Add as much ram/disk as you can because once people start using vms they add 
more all the time and those two constraints are usually the issue.

The downside is that some administration or failures may mean powering off 
or losing all servers. To deal with this people use redundant hosts with hot 
swappable disk and p/s. The bigger ones put in shared storage and a 
redundant host and do live migration of running vms.



----- Original Message -----
From: kwlug-disc-bounces at kwlug.org <kwlug-disc-bounces at kwlug.org>
To: kwlug-disc at kwlug.org <kwlug-disc at kwlug.org>
Sent: Fri Nov 05 23:58:38 2010
Subject: [kwlug-disc] Virtualization allocation?


I am (foolishly) starting to think about virtualization.

Right now I have a non-virtualized Linux server that serves up several
roles:
- Trouble ticket system
- Nagios alerts
- Cacti graphs
- Certificate generation
- Internal mail server
- Lightweight FTP repository
- ...


Naturally, this is one of those servers that started out being a
development/staging/proof-of-concept machine and grew. The trouble is
that the machine works pretty well even though it has all of these
roles.

As I dabble my pinky-toe into virtualization, I am wondering whether it
makes much sense to split up a box that offers these 5-10 services into
5-10 distinct VMs.

I understand that recovering from a bad hardware failure will be
problematic. I understand that (in principle, at least) splitting up
roles increases security. But I am having a hard time justifying why I
would want to take an old Pentium III that works well and turn it
into several slices of some much more powerful (and expensive!)
machine.

I would not be splitting up this box right away even if I decided to,
but I am at the stage where I am thinking about what kind of hardware
I am going to need to support virtualized servers in production.

Convince me?

For those of you who do Linux server virtualization: how do you split
up the load?

- Paul

-- 
http://pnijjar.freeshell.org


_______________________________________________
kwlug-disc_kwlug.org mailing list
kwlug-disc_kwlug.org at kwlug.org
http://astoria.ccjclearline.com/mailman/listinfo/kwlug-disc_kwlug.org