response 58 of 60:

Mar 2 07:00 UTC 2010

  
actually, changing the list can be done by anyone -on- hte list except
that i got hte list changed to include me and remmers before i was
on teh list. it could have been done earlier & faster a few months
ago but i didn;t wnat to stomp on toes.
  
if there is a panic hands-on-rqueired situation it;s true that we aer
subject to proide's opne-hours. but less so than we were in ken's wharehouse
since provide is open weekends. having a site like lmaster's basemsnt
or the access of the pumpkin isn't, imo, all that critical altohogh it
was nice.
  
bsesides, there are more of us to assist now.  awareness can./could go
back to seleep again but that;s remote for the near fture, imo.

response 60 of 60:

Mar 6 14:48 UTC 2010

It wouldn't be difficult to mirror some of Grex's facilities -
e.g. bbs and the website - such that if Grex's "main machine"
became unavailable, access could be switched over to the mirror.
Grex's storage and computing requirements are so modest that I
think that mirroring "in the cloud" is a pretty inexpensive
proposition.

Just mirroring the data is trivially easy.  As an experiment, I
used rsync to copy all of the Agora conferences (dating back to
1991) to a FreeBSD virtual machine at my disposal.  Creating the
initial mirror took 36 minutes of elapsed time - about 380 mb of
data.  The mirror can be kept in sync by running rsync as a
cronjob at frequent intervals.  Since rsync copies only the
changes since the last run, resource usage for keeping the mirror
synchronized would be quite low.  Now, if you have backtalk running
on the mirror and also keep users' participation files sync'd, 
people could participate in bbs on the mirror if the main
machine was unavailable.

Turning that into a production system would take a bit of thought
but shouldn't be too difficult.  But anyway, I think it would
behoove Grex to implement some redundancy so that it doesn't rely
on the availability of just one machine.  It wouldn't be expensive
to do this.