Azure redundancy for Virtual Servers (Linux)

Think what you should look at is Azure managed images.. should allow you to create a point in time image of the VM, to redeploy from later, if I understand it correct.. I am on AWS myself, so not intimately familiar with Azure..
 
WAslayer said:
Think what you should look at is Azure managed images.. should allow you to create a point in time image of the VM, to redeploy from later, if I understand it correct.. I am on AWS myself, so not intimately familiar with Azure..
Click to expand...

This can work.
 
They have Azure Backup - it wouldn't be my first choice for protecting on premise servers but it is probably OK for backing up Azure VM's and was relatively cost effective last time I Iooked at it.
 
Eric said:
What options are there for redundancy on Azure virtual servers without 'rolling your own'?

I see redundant storage options, I assume that amounts to data backup only which would require me to fire up a new server, install the OS and components and then attach the backup storage?

What I'm hoping for is something that would allow me to take snapshots of the OS disk and allow me to quickly deploy a new VPS in the event of disaster without having to go through the whole installation and configuration process (or have to identify and backup all configuration files).



What do you use or recommend?
Click to expand...
You're looking for the Availability option where you can sync your VMs to another Availability Zone (AZ). Bear in mind the costs, but the failover is seamless. I have done this for a client and it works very well. Pretty easy to setup too.
You could, for instance, host in South Africa North, and sync to South Africa West

Availability options - Azure Virtual Machines | Microsoft Docs
 
Ja, but something else to look at is Azure Site Recovery. Sorry, ignore the Availability Sets for now Eric. This is what I did. Got distracted by Availability sets and meant this. My apologies.

This will allow you to DR between zones in the same region. You can set up policies to failover automatically, and it allows you to do a test failover as well. Works brilliantly!

Azure Site Recovery | Microsoft Azure
 
Eric said:
Awesome, will take a look at that! Cheers :)
Click to expand...
Sorry again for the incorrect info...but Azure SR works exceptionally well. I don't see South Africa West (Cape Town) for some reason in the list of Azure Zones, so you MIGHT have to failover to another region (say UK). Just check out where you can and cannot fail over too. The support documentation is very good.
 
ZA_Chamber said:
Sorry again for the incorrect info...but Azure SR works exceptionally well. I don't see South Africa West (Cape Town) for some reason in the list of Azure Zones, so you MIGHT have to failover to another region (say UK). Just check out where you can and cannot fail over too. The support documentation is very good.
Click to expand...
Cape Town is the availability partner for Joburg,but the DC is only "selectable" for manual failovers on the premium option. Otherwise on the basic option it'll be failed over by Azure for a Joburg DC outage last time I checked

I have a few client sites where we use it on regions like UK South/West. The replicas are normally within 5minutes for changes. Manual failover takes about 2.5min excluding boot times
 
PsyWulf said:
To be fair,the recovery time for Geo-redundant backups is under 3 hours from experience,so depends on the RPO/RTO
Click to expand...

IIRC with Azure Backup you also get the option to have a couple of snapshot days alongside the data retention options, which if I'm not mistaken is just access to faster storage/recovery. So unless OP wants a seamless cutover and can't afford a bit of downtime then I'd still consider it?

And then he has both backup (compliance of 7 years retention if he sets it up so) and redundancy (GRS backup) checkboxes done.
 
Last edited:
Bobbin said:
IIRC with Azure Backup you also get the option to have a couple of snapshot days alongside the data retention options, which if I'm not mistaken is just access to faster storage/recovery. So unless OP wants a seamless cutover and can't afford a bit of downtime then I'd still consider it?

And then he has both backup (compliance of 7 years retention if he sets it up so) and redundancy checkboxes done.
Click to expand...
Yeah that was kinda the point I was making. Depends how quickly they need to fail over ;)
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X