SheepStealer
asked on
vmware fault tolerance site to site redundancy
Hi,
I have been tasked with setting up vmware site to site failover for various windows machines
I have a few questions that I would be so grateful if the community here pitched in on.
I assume the feature I am wanting to use is the fault tolerant VM in vsphere
what will I need to do at the front end to ensure internet users are passed smoothly between a failed site/server and its fault tolerant copy at another site?
does anyone have experience too of the bandwidth required between sites?
Also will Windows Clustering have similar network needs if anyone knows?
Many thanks
Rollo
I have been tasked with setting up vmware site to site failover for various windows machines
I have a few questions that I would be so grateful if the community here pitched in on.
I assume the feature I am wanting to use is the fault tolerant VM in vsphere
what will I need to do at the front end to ensure internet users are passed smoothly between a failed site/server and its fault tolerant copy at another site?
does anyone have experience too of the bandwidth required between sites?
Also will Windows Clustering have similar network needs if anyone knows?
Many thanks
Rollo
Firstly, how much downtime can you afford ?
30 seconds, 1 minutes, 2 minutes, 30 minutes?
30 seconds, 1 minutes, 2 minutes, 30 minutes?
ASKER
Hi,
wow thanks for your quick responses.
I guess we would be after is it where both sites are active.
then if there is a failure i guess other servers at the working site can come up to take the load.
I thought with fault tolerance it could be seamless ? that would be ideal of course.
I just can imagine how it can work say for someone VPNd in for the endpoint to basically move to another site without the tunnel crashing.
I thought DNS records were cached all over and updated kind of randomly.
is this what professional load balancers do. Gosh but they are expensive is there a cheaper solution?
I was hoping everything was going to be built with just vmware kit we have 8 CPU of essentials plus. I just don't understand I suppose how the front end works?
surely the session to the server has to be abstracted and handled by a vSwitch somehow to be smoothly moved.
I have used Veeam it was great I am sure we will be using that for replication offsite.
thanks again.
wow thanks for your quick responses.
I guess we would be after is it where both sites are active.
then if there is a failure i guess other servers at the working site can come up to take the load.
I thought with fault tolerance it could be seamless ? that would be ideal of course.
I just can imagine how it can work say for someone VPNd in for the endpoint to basically move to another site without the tunnel crashing.
I thought DNS records were cached all over and updated kind of randomly.
is this what professional load balancers do. Gosh but they are expensive is there a cheaper solution?
I was hoping everything was going to be built with just vmware kit we have 8 CPU of essentials plus. I just don't understand I suppose how the front end works?
surely the session to the server has to be abstracted and handled by a vSwitch somehow to be smoothly moved.
I have used Veeam it was great I am sure we will be using that for replication offsite.
thanks again.
an active-active ESX clustered environment will be a very expensive solution . how your sites are connected ? MPLS ? VPN ?
ASKER
HI akhaliighi,
I think I am not sure.
I know MPLS is the gold standard for flexibilty.
what will make it expensive can you advise?
many thanks
I think I am not sure.
I know MPLS is the gold standard for flexibilty.
what will make it expensive can you advise?
many thanks
So you state both sites active, e.g. 0 seconds failure?
ASKER
HI Hancocka,
is this possible? it would surely be ideal.
it is just how to accomplish this?
will it need to be virtual IP to be abstracted between the sites?
regards
is this possible? it would surely be ideal.
it is just how to accomplish this?
will it need to be virtual IP to be abstracted between the sites?
regards
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Are you looking for an active-active solution or active-passive ?
in a nutshell , if this is ONLY for disaster recovery ( active-passive) ; you will need a solution to replicate your VMs from primary site to secondary . depends on your business requirements , you may want to replicate VMS every night or every hour or etc .
to replicate VMs ; you can use SAN-to-SAN replications , solutions like DATACORE or you can use 3rd party applications that run replications between ESX servers . a famous one is veeam.
to transfer end users from primary to secondary in case of a fail-over you can use DNS .