3 ms·

> The costs would be absolutely fixed and known. Until the server breaks and you have to drive over in the middle of the night and try to replace it but the on

> The costs would be absolutely fixed and known. Until the server breaks and you have to drive over in the middle of the night and try to replace it but the only available server right now is a shitty one and oh shit only half the backups work cause the onsite backups are fried too etc etc etc. There's many good arguments against high-level BaaS such as Firebase but I'm not sure that "colo is cheaper" is one.

walrus01 6y ago
fine, buy two identical ones. and set up proper backups. or even make the backup a hot-spare. if it's a test/development system that's meant to possibly break, you shouldn't be driving anywhere at 0300 in the morning anyways.
- wrkronmiller 6y ago
  That implies you are going to be running prod in the cloud. Unless you're developing against purely synthetic data, the data transfer costs are potentially astronomical.
daneel_w 6y ago
It's anecdotal, but I'm convinced I'm not alone here: we've had more Amazon-related failures/outages in 3 years with AWS than we had in 4 years of colo before heading to the cloud because of the exact fear you described. Even a cloud setup needs good management and contingency planning, and in absence of such it can fail just as hard as a colo setup.
- walrus01 6y ago
  https://www.bbc.com/news/technology-55087054 https://www.bbc.com/news/technology-55087054 https://www.seattletimes.com/business/amazon/amazon-web-services-outage-hits-cloud-customers/ https://www.seattletimes.com/business/amazon/amazon-web-serv...
z3t4 6y ago
Just because you use "the cloud" doesn't mean you don't need backups. "the cloud" also have downtime and other failure. When deploying to the cloud you have to factor in the cost for moving to another provider if/when it will be needed.
senko 6y ago
> I'm not sure that "colo is cheaper" is one. It absolutely is (cheaper, and a good argument). As an example: we're in the process of switching from Digital Ocean to Hetzner for a project, that will increase infrastructure performance (roughly memory/cpu/storage) by 4x and decrease costs by 4x. And no driving to the colo center is neccessary, as it's their dedicated server, so their on-site engineers do the hardware replacement. Also, if you are not okay with your site being down for a few hours, you can always buy two, like you would with a sensible cloud setup. It'd still come up way cheaper (+ get more perf if you can do load balancing for your usage). Also, I don't look at it from "colo is cheaper" point of view. To me, it's "I can have several times more performance and hire a full time sysadmin to worry about it, for the same price".
- Axsuul 6y ago
  Can anyone recommend a similar provider but for North America?
  - 1337shadow 6y ago
    Sure thing https://oneprovider.com/dedicated-servers-in-north-america https://oneprovider.com/dedicated-servers-in-north-america

walrus01 6y ago

fine, buy two identical ones. and set up proper backups. or even make the backup a hot-spare. if it's a test/development system that's meant to possibly break, you shouldn't be driving anywhere at 0300 in the morning anyways.

wrkronmiller 6y ago
That implies you are going to be running prod in the cloud. Unless you're developing against purely synthetic data, the data transfer costs are potentially astronomical.

daneel_w 6y ago

It's anecdotal, but I'm convinced I'm not alone here: we've had more Amazon-related failures/outages in 3 years with AWS than we had in 4 years of colo before heading to the cloud because of the exact fear you described. Even a cloud setup needs good management and contingency planning, and in absence of such it can fail just as hard as a colo setup.

walrus01 6y ago
https://www.bbc.com/news/technology-55087054 https://www.bbc.com/news/technology-55087054 https://www.seattletimes.com/business/amazon/amazon-web-services-outage-hits-cloud-customers/ https://www.seattletimes.com/business/amazon/amazon-web-serv...

z3t4 6y ago

Just because you use "the cloud" doesn't mean you don't need backups. "the cloud" also have downtime and other failure. When deploying to the cloud you have to factor in the cost for moving to another provider if/when it will be needed.

senko 6y ago

> I'm not sure that "colo is cheaper" is one. It absolutely is (cheaper, and a good argument). As an example: we're in the process of switching from Digital Ocean to Hetzner for a project, that will increase infrastructure performance (roughly memory/cpu/storage) by 4x and decrease costs by 4x. And no driving to the colo center is neccessary, as it's their dedicated server, so their on-site engineers do the hardware replacement. Also, if you are not okay with your site being down for a few hours, you can always buy two, like you would with a sensible cloud setup. It'd still come up way cheaper (+ get more perf if you can do load balancing for your usage). Also, I don't look at it from "colo is cheaper" point of view. To me, it's "I can have several times more performance and hire a full time sysadmin to worry about it, for the same price".

Axsuul 6y ago
Can anyone recommend a similar provider but for North America?
- 1337shadow 6y ago
  Sure thing https://oneprovider.com/dedicated-servers-in-north-america https://oneprovider.com/dedicated-servers-in-north-america