Re: [cobalt-users] How are people handling fault tolerence?

[Jeff Lasman <jblists@xxxxxxxxxxxxx>]

Chip wrote:

> For those of you that are hosting 100 sites or less, how do you handle
> redundancy and/or failure?  I see several possibilities:

While I like Kris' reply, I don't think most of us are willing to go
that route, especially for only a few sites (you can define a "few" as
you wish <smile>).

So I see two options...

> 1. Keep good backups and hope the Raq doesn't go down, then hustle to get
> another one if it does. (Not a very good choice, IMHO, but a lot of folks
> seem to be doing this)

Good if, as Kris recommends, you have another drive ready and have good
backups, AND have fast physical access to your colocation point, or a
service contract that includes restoring a down server.

> 3.  Buy 2 Raq servers, mirror information between them and locate in same NOC

or...

> 4. Same as above, but locate them in different ISPs with different NOCs.

I'd go with one of the above, depending on how bulletproof your
colocation provider is.  None of these options are, as Kris points out,
for true high-availability hosting, Cobalt really doesn't make a
solution for that.

Since you can usually get a break on the second RaQ at the same
location, share the bandwidth, AND avoid bandwidth for nightly backup of
changed files.

I'd NOT try for load-balancing, and ONLY use the second RaQ for backup. 
And I'd do that through custom-written scripts running from cron, and
NOT depend on Cobalt's backup, since so many people have had a lot of
problems with it.

You'd probably still need quick physical access to your server, or a
good service contract with a colo company with 24/7 remote-hands
service.

And, for you purists out there, I know this is nowhere near a
bullet-proof solution.

But it's better than what most of us do now.

Jeff
-- 
Jeff Lasman <jblists@xxxxxxxxxxxxx>
nobaloney.net
P. O. Box 52672
Riverside, CA  92517
voice: (909) 787-8589  *  fax: (909) 782-0205