Sucks to hear that, but good to see an oldtimer posting again. :)

all but one of our servers is now up. Of course its the one that houses OHC :) hehe

Anyone who says power is an easy thing to do... is an idiot.

Generators fail, ups explode (they really do), copper wire explodes from metal fatigue. not to sound crazy... but power failure is a sign of success... why? because people are using yoru service!

what about proper maintenance? Still happens.

While I am prone to giggling over the downtime as well... and despite really not liking them at all for our own personal reasons.... shit happens.


The bottomline is this:
unless you spent a mint on datacenter, your shit is going down at some point.
not going down will prolly cost you double at least. :2 cents:

My bad

I always looked at Choopa as a cheap host cuz they've spammed their services on the boards.

This is why we own our own data center now because when you colocate in someone's else building it becomes a nightmare. Anyways best of luck to choopa and their customers.

To add even more wobble, the replacement router or some related part caused maximum throughput to drop to 0.5Mbps

This is NOT a good start...

Tough break
 

Your guys and gals unfortunate situation caused me to ask my web guy, could that happen to any of our sites hosted at Dedico?

I now have the piece of mind of knowing that we are being hosted in a sas70 type-II facility with 100% power and cooling uptime SLA.

Can someone translate that into english for me? :winkwink:

English: No datacenter is immune from downtime.

100% SLA sounds good, but it's not much use getting credited a couple of hundred bucks in hosting fees when your immediate and future losses because of any incident could work out to be significantly more.

Depending on your sites and setup it may be possible to have a "warm spare" server or VPS image sitting at another host, ready for activation at short notice.

Ive had my Equinix Cage for 7 Years and still going Strong, never had any kind of power issue. I pay a premium for this and im more than glad to pay the extra to have peace of mind.

The Halsey building has always had power issues, so its no surprise about Choopa going down

Setting up hot spares is certainly an option but most won't do it once they see the expenses involved. If your sites are BW intensive then your costs may no be as expensive as your main site but if BW is minimal then the costs may actually exceed the cost of the primary hosting.

Don't jinx yourself. You are obviously leaving in a dream world as there have been plenty of Equinix downtimes. Maybe not your specific datacenter, but what about all the other people paying a premium in their other ones?

Another reason to spread servers and stuff across at least 2 datacenters :-)

Interesting thought. Thanks.

Here's one such service, I'm not sure if it's dedicated or VPS --> https://www.bettercgi.com/clonebox/

If you have the budget and ability to run a hot spare then it's probably worth just doing full time load balancing...

Ever since the last major power outage in the US, not only do we use multiple data centers but the data centers must be on different power grids.

Whats even better is finding a host that has datacenters in completely different states, running completely independent from each other -- YET providing a secure, internal, FREE link between both locations.For example, datacenter say in Los Angeles and in Dallas. The clients then fully benefit from the advantages of having diverse DC locations, yet the machines interact with each other as if they were in the same rack next to each other.

Did I mention that we offer this? Its called the PRIVATE NETWORK. Not only does it interconnect all servers in our Los Angeles datacenters to each other on a dedicated routed internal network, but it also connects to our Dallas datacenter.

It will also connect to our THIRD (Chicago or East Coast, TBD) and FOURTH (Europe) locations soon too. So you can diversify your machines in 4 geographically separate locations (West USA, Central USA, East USA, and Europe) and still have them interact as if they were all in the same server cabinet.

Thats the end of my spam for now. :)

Do you know how do deal with memcached servers? Just like Facebook does it??? reference: http://www.facebook.com/note.php?note_id=39391378919
more reference: http://github.com/fbmarc/facebook-memcached

Im already using it out of a NOC from NYC. I need to spread my wings out to the West Coast and beyond.

If so ICQ me with an memcached server IP and port to test it out.

If you are a rep don't bother, I need someone who knows what they are talking about. Most reps don't know shit, good reps point to the people who know the shit.