[HCoop-Discuss] Our ideal architecture?

[Adam Chlipala]

Davor Ocelic wrote:
> Openvz has very detailed resource usage limits, which are also 
> conveniently displayed as a table in one of the /proc files, where
> a person can easily see existing containers, their limits, current
> usage and number of times the limits were overstepped.
>
> Here's partial output from that file, /proc/user_beancounters
> ("fitted" in 80 columns without wordwrap for your viewing pleasure):
>
>  uid  resource       held maxheld   barrier                limit  failcnt
> 104:  kmemsize     857224 9942062  12752512             12752512        0
>       lockedpages       0       0       256                  256        0
>   

Does "uid" mean "container ID" here? Are containers somehow conflated 
with users?

> For example, if we were to give out VMs to members, I don't see ANY
> advantage that HCoop would have compared to any other VM provider.

Well, there are the advantages of democratic organization, but those 
don't seem to motivate that large a fraction of our current membership, 
so it probably also wouldn't matter much to potential VPS users.

One big one, though, could be the ability to mount HCoop AFS volumes. A 
fancy-pants distributed file system maintained by someone else is 
nothing to shrug off. In this scenario, a VM user could also rely on us 
to back up all AFS volumes automatically and in a reasonable way.

There are a slew of other possible benefits of being near shared 
machines, like being able to set up monitoring, development systems, 
etc., in the same data center (with fast/free transfer), without needing 
to manage extra machines.

I do strongly agree that we should ignore support for any per-member 
VPSes until after we are happy with services along our current model, 
though.