cPanel High Availability

[hostmedic]

cPanel High Availability

What are the suggested methods to ensure uptime for a server?

I am wondering what others out in the market are doing...

Raid ? Clustering ?

[tweakservers]

there's some thread discussion on http://forums.cpanel.net/showthread.php?t=34925 which you might want to have a look.

[hostmedic]

thanks

I am actually part of that discussion...

What I was thinking about was looking to see how others have setup servers...


Raid / Mirror or ...?
MySQL Master -Master ?
etc ...

[Davec]

Sorry i am not good in this. But I am wondering why there is noone / companies going to setup a services which could provided such services. A business model like the rvskin / Installatron running as a third party at the same time with a close relationship with Cpanel.

Of course, this company or engineer must prove the success with such high availability. With the amount of works and experience already placed in the past and continue provide the technical support and update services. An annual license fee payment is required for those who would use such services.

What do you guy thinks?

[n3tph4t]

Sorry i am not good in this. But I am wondering why there is noone / companies going to setup a services which could provided such services. A business model like the rvskin / Installatron running as a third party at the same time with a close relationship with Cpanel.

Of course, this company or engineer must prove the success with such high availability. With the amount of works and experience already placed in the past and continue provide the technical support and update services. An annual license fee payment is required for those who would use such services.

What do you guy thinks?

..because the business model doesn't add up. Whilst these services are a requirement for some, it doesn't scale and the market is limited.


For both capacity & availability management:
Raid(/SAN)
LVS(/heartbeat loadbalancing)
Mirrored SQL instances
Mirrored DNS
Warm standby equipment
Redundant connectivity
Dual power feeds in the rack(s)
UPS with backup generator
Dual-sited equipment across a geographical area (this would make redundant connectivity easier but routing would be more difficult depending on your scale).
Fire suppression systems (in almost all datacentres anyway)
Change management & configuration management processes
Limited access physically and logically

--the last two because IMHO the biggest trick to availability is normally change!


At a guess