Ovh sbg tzunami

[leonep]

Hi guys,
I am 1 fortune user with server in Sgb3. The good news is my server is safe.
OVh CEO has told us on Twitter we get back access on 19.
The dirty part of this story is that my backup on Roubaix is not accessible because I can only access with server ip address.this for security and support cannot help.
Now if I want migrate I cannot migrate because I do not have data.
After this apocalipse I want to change my plan and find a way to create high availability services.
I want 2 server geo separated working in cluster with same data always in sincro. Most are web and mail.
How to realize this?
Thanks

 

[cPRex]

Hey there! Currently cPanel does not have any failover support like that available. It's definitely something we're working toward with the release of tools like Mail Node, but there isn't a way to have a live backup server with cpanel at this time. You would need to utilize custom third-party tools in order to get that behavior you're looking for.

 

[leonep]

Thanks,
I start to search something that is good for me.
What is the most used 3rdparty tools used by little company like mine?
unfortunately there is also to say that people do not want to spend a lot of money, they think that the cloud is immortal !
we must balance all
thanks for listen to me , i am really frustrated , i hope you understand !
regards

 

[cPRex]

Since this isn't something we test on our end I don't have any official recommendations to provide. I'm going to move this post over to the hyperscalers section of the forum for increased visibility, so another user could comment with their experiences.

 

[vacancy]

You can access the backup area from outside by giving the ip address permission from the ovh backup settings.

 

[leonep]

It is possible to access only with ovh failover ip ....
so i must purchase anotther server only for donwload backups...
do you know if it is possible with a vps?
thanks

 

[vacancy]

Yes, that makes sense. You can get your backups there.

 

[leonep]

@vacancy you are better of ovh support
i will update you!
thanks

 

[gnusys]

While this is unfortunate. I have seen in my career of over 15 years 2 incidents where a transformer blew and put out an entire DC down for over a week and now a fire destroyed many servers and put out access to many others. You need a multi-datacenter cluster to prevent such incidents. Although rare, accidents/natural calamities, etc are a risk that must be considered for your online business and data. You can search in the cPanel App catalog: cPanel App Catalog :: Third-Party Applications for cPanel & WHM for tools/software that let you deploy multi-data center, high available and redundant infrastructure.

 

[leonep]

i saved my data , setup a new server and all is fine now. but i was down 8 days...
i need to find a better strategy
thanks

 

[HostXNow_Chris]

i saved my data , setup a new server and all is fine now. but i was down 8 days...
i need to find a better strategy
thanks

Now use Acronis to backup/restore data faster in future

 

[leonep]

acronis backup got high aviability feature?

 

[HostXNow_Chris]

acronis backup got high aviability feature?

Usually, production services should be high availability, and backup is just that a backup.

But Acronis allows you to run your own setup and use different storage using Acronis Cyber Infrastructure. Still, if you use Acronis Backup Cloud, then they take care of that for you and no doubt duplicate the data between DCs.

https://www.acronis.com/en-gb/data-centers/

High data availability
Ensure that data required for your business operability has one of the highest levels of availability, leveraging data centres that provide redundant HVAC, network and UPS systems.

Is it also possible to copy data from one DC to another with Acronis when using Cloud storage? That said, Acronis isn't using the same setup OVH did and doubt their DC would go up the way OVH's did. OVH did not hide that SBG2 was an old setup and was supposed to be decommissioned because they left it a bit late, unfortunately.

 

[monarobase]

You can do high availability with VMs. Full automatic failover with NFS or CEPH will cause lower disk I/O and higher disk response times as well as I/O wait issues when there is a network issue between storage and compute servers.

We prefer to keep I/O performance and stability and instead sync the VM between two servers in two datacenters every minute. In case of a total server loss we can just manually restart the server on the other node.

This does require more knowledge and more hardware but it much faster to recover a full server this way and removes the issue of not being able to replace the server quickly because you already have a physical node ready instead of waiting for the DC to replace the failed server pr order a new server if they cannot.