Orchestrate and control many web sites
-
@black3dynamite said in Orchestrate and control many web sites:
I'm curious, do you have an example of a PAAS that SaltStack might not work well or at all?
There's no system in which to install the Salt-Minion client I'm guessing.
Does the PAAS instances give SSH access?
-
@Obsolesce said in Orchestrate and control many web sites:
There's no system in which to install the Salt-Minion client I'm guessing.
In that case, if SSH is available, Ansible would be my choice.
-
@black3dynamite said in Orchestrate and control many web sites:
@Obsolesce said in Orchestrate and control many web sites:
There's no system in which to install the Salt-Minion client I'm guessing.
In that case, if SSH is available, Ansible would be my choice.
He said that SSH is not available everywhere.
-
@Emad-R said in Orchestrate and control many web sites:
Why not move them, well Business, shared hosting is cheaper with plenty of resources like disk space. and things are already like this.
Not really. That's only cheaper when you are so small as to not have any IaaS at all. Even NTG is big enough on internal web sites to have shared hosting be more expensive.
-
Not sure in a case like this that any universal solution can really exist, unless you wrote your own specifically to address the limitations in question.
-
@Obsolesce said in Orchestrate and control many web sites:
@black3dynamite said in Orchestrate and control many web sites:
I'm curious, do you have an example of a PAAS that SaltStack might not work well or at all?
There's no system in which to install the Salt-Minion client I'm guessing.
Does the PAAS instances give SSH access?
-
@scottalanmiller said in Orchestrate and control many web sites:
@Emad-R said in Orchestrate and control many web sites:
Why not move them, well Business, shared hosting is cheaper with plenty of resources like disk space. and things are already like this.
Not really. That's only cheaper when you are so small as to not have any IaaS at all. Even NTG is big enough on internal web sites to have shared hosting be more expensive.
Check this price 2.95mo*
https://www.tmdhosting.com/shared-hosting.html
Unlimited SSD Space Unlimited Bandwidth 1 Website Hosted Free Domain Premium Support Free cPanel NGINX Web Server Basic Caching Let's Encrypt SSL 60 days Money back guaranteeWe have sites that are big in files so Vultr or DO will make things alot expensive
-
@Emad-R Exactly, that's more than running your own. $2.95 PER WEBSITE is quite expensive. Run your own and you can do it for $.35 per site!
-
I will do some calculations and get back to you on that one.
-
@Emad-R said in Orchestrate and control many web sites:
I will do some calculations and get back to you on that one.
I run a commercial web host. All of the cost is in the site labour, not the hosting per se.
-
@scottalanmiller said in Orchestrate and control many web sites:
@Emad-R Exactly, that's more than running your own. $2.95 PER WEBSITE is quite expensive. Run your own and you can do it for $.35 per site!
No you cannot. Not for a single site or even a dozen.
$5 vultr instance / 12 sites comes in at $0.42
That doesn't count the labor to maintain, patch, make resiliant (excluding content related development, etc).
While certainly not expensive, it is also certainly not $0.35. -
@JaredBusch said in Orchestrate and control many web sites:
@scottalanmiller said in Orchestrate and control many web sites:
@Emad-R Exactly, that's more than running your own. $2.95 PER WEBSITE is quite expensive. Run your own and you can do it for $.35 per site!
No you cannot. Not for a single site or even a dozen.
$5 vultr instance / 12 sites comes in at $0.42
That doesn't count the labor to maintain, patch, make resiliant (excluding content related development, etc).
While certainly not expensive, it is also certainly not $0.35.Ah, but don't forget, when we did a labour study we found that dealing with your own install had lower labour than when dealing with a shared instance because you don't get raw access. So we lowered labour.
So my numbers were obviously without labour, but so were the shared numbers. If you include labour both go up, but the shared goes up more from what we found in testing.
