6 ms·

One issue with this is the idea of competition from physical hardware. At some price point, if the prices are too high, it will be cheaper to hire extra people

One issue with this is the idea of competition from physical hardware. At some price point, if the prices are too high, it will be cheaper to hire extra people, and get a cage in a few datacenters and some physical hardware.

bookmarkacc 10y ago
The problem is people who arent big enough to have a whole data center. Or dont have the upfront capital even if its benefital in the longterm. They will still rely on Amazon.
- Johnny555 10y ago
  You don't need to own a whole datacenter to own machines located in a datacenter, my company keeps machines in a 4 rack cage in a datacenter. (and plans to empty that cage by the time our contract runs out in September to finish the migration to AWS). Our AWS bill is just a bit higher than what we were paying for the cage rental + internet connectivity alone.
  - nopzor 10y ago
    That doesn't sound possible. Are you saying your raw bill for _just_ colo (ie. space and power) and bandwidth (ie. IP transit) was less than your _entire_ AWS bill? I've never heard of anyone take such an extreme stance.
    - Johnny555 10y ago
      The 4 rack cage plus extra power plus redundant network at a Tier I datacenter costs us close to $6K/month (including some block of remote hands hours). Even at our peak we didn't have the racks full, allowed room for expansion. Note that this cost includes only coloc costs, not server or network leases/purchases. We have very bursty traffic and run our baseload of ~ 30 severs (mostly m4.larges, with a few smaller utility servers as well as larger servers for back end services) most of the time, but when we need to scale, we ramp up to about 2 - 4 times that, but usually only for a few hours at a time. We also do some batch processing that takes a few high-memory servers for about a day each week. We monitor costs pretty closely and our monthly AWS spend is still less than the coloc fees, and we are much more scalable now and have multi-datacenter (AZ) redundancy, and by the end of the year, we will have a cold spare in a second region (which will cost us very little since it will be mostly powered off) We no longer need to vastly over-provision servers to handle our peak load and batch processing needs.

bookmarkacc 10y ago

The problem is people who arent big enough to have a whole data center. Or dont have the upfront capital even if its benefital in the longterm. They will still rely on Amazon.

Johnny555 10y ago
You don't need to own a whole datacenter to own machines located in a datacenter, my company keeps machines in a 4 rack cage in a datacenter. (and plans to empty that cage by the time our contract runs out in September to finish the migration to AWS). Our AWS bill is just a bit higher than what we were paying for the cage rental + internet connectivity alone.
- nopzor 10y ago
  That doesn't sound possible. Are you saying your raw bill for _just_ colo (ie. space and power) and bandwidth (ie. IP transit) was less than your _entire_ AWS bill? I've never heard of anyone take such an extreme stance.
  - Johnny555 10y ago
    The 4 rack cage plus extra power plus redundant network at a Tier I datacenter costs us close to $6K/month (including some block of remote hands hours). Even at our peak we didn't have the racks full, allowed room for expansion. Note that this cost includes only coloc costs, not server or network leases/purchases. We have very bursty traffic and run our baseload of ~ 30 severs (mostly m4.larges, with a few smaller utility servers as well as larger servers for back end services) most of the time, but when we need to scale, we ramp up to about 2 - 4 times that, but usually only for a few hours at a time. We also do some batch processing that takes a few high-memory servers for about a day each week. We monitor costs pretty closely and our monthly AWS spend is still less than the coloc fees, and we are much more scalable now and have multi-datacenter (AZ) redundancy, and by the end of the year, we will have a cold spare in a second region (which will cost us very little since it will be mostly powered off) We no longer need to vastly over-provision servers to handle our peak load and batch processing needs.