Server Redundancy & Ghosting Domain Controllers.
- Thread starter BoomAM
- Start date
Sorry forgot completley about this post.
Where this happened was in a building society. As you can imagine any downtime what-so-ever looses the company money.
Clustering isn't necessarily the best solution however it is the only one that fits in with the OP's requirements. As previous posters have said having another server acting as a DC is dependant upon what you replicate and if one fails then you may need to change settings to get the other server working as the PDC.
We had an environment where we basically had 4 DC's for each site (3 sites) a PDC and a SDC cluster (so two serperate clusters) the reason for this was that if any of the servers went down we could pull them out and there was no down time and when your bits are on the line its good to have that kind of redundancy. We also did not use MS clustering, the name escapes me at the moment of the product we used - I'm sure it'll come to me later.
As I said it's not ideal for everyone but it's good to experiment with.
M.
Where this happened was in a building society. As you can imagine any downtime what-so-ever looses the company money.
Clustering isn't necessarily the best solution however it is the only one that fits in with the OP's requirements. As previous posters have said having another server acting as a DC is dependant upon what you replicate and if one fails then you may need to change settings to get the other server working as the PDC.
We had an environment where we basically had 4 DC's for each site (3 sites) a PDC and a SDC cluster (so two serperate clusters) the reason for this was that if any of the servers went down we could pull them out and there was no down time and when your bits are on the line its good to have that kind of redundancy. We also did not use MS clustering, the name escapes me at the moment of the product we used - I'm sure it'll come to me later.
As I said it's not ideal for everyone but it's good to experiment with.
M.
No where does he ask for the 'easiest way' to do it. He just asks how easy it is.
Also although it might not be easiest way to do it, its probably the best for redundancy thats why I and few others were recommending it. Never said its the only way to do it but he was asking for advice and I'm not the sort of person to look for the easiest way out as it normally comes back to bite you in the end.
Also although it might not be easiest way to do it, its probably the best for redundancy thats why I and few others were recommending it. Never said its the only way to do it but he was asking for advice and I'm not the sort of person to look for the easiest way out as it normally comes back to bite you in the end.
It'snot about having the easiest way out, it's about giving the best advice, you've not done that.
Let me reword then. If someone asks how 'easy' something is, would you offer them a simple industry standard solution, or not?
http://technet2.microsoft.com/windo...8de9-4ea5-9c66-33605823bbc71033.mspx?mfr=true
Read the above, then post back if you think clustering might be the best solution for the OP. He has one DC, and needs redundancy, what does that tell you, to install a cluster?
Let me reword then. If someone asks how 'easy' something is, would you offer them a simple industry standard solution, or not?
http://technet2.microsoft.com/windo...8de9-4ea5-9c66-33605823bbc71033.mspx?mfr=true
Read the above, then post back if you think clustering might be the best solution for the OP. He has one DC, and needs redundancy, what does that tell you, to install a cluster?
If you want redudancy yes. Redudancy to me means that having as much uptime as possible is important.
If he just wants to be able to recover from a DC failure faster then I would recommend a second DC and share the roles between them
If he just wants to be able to recover from a DC failure faster then I would recommend a second DC and share the roles between them
A second server isnt really an option at the moment, although in the next 12 months, we might be replacing an equal server on our second network, that we could use on the primary network at a second DC.
The option i want is ghosting or imaging the DC once a month to a NAS.
The option i want is ghosting or imaging the DC once a month to a NAS.
mr.bond said:
Well if you're saying it's about giving the best advice then surely the most up time with the least configuration (after the initial configuration) is surely sound advice. I don't know the OP's technical level or the technical competance of the staff that maybe employed by the company.
As I said in my previous posts there are lots of software out there to make clustering easier. Some can actually do it in a few clicks (of course you need a few spare IP's).
Clustering OR having a second DC is the industry standard. The entire forums are about giving people different ideas - it's now up to the OP to investigate the solutions offered and choose a way forward - hopefully he should be more informed than he was when he wrote the post - and thats all I'm trying to do - give him options!
M.
Oh and with Ghost / Acronis - please note that you will make an exact image of the server - so if you restore it to another server (hardware must be exactly the same otherwise you will get issues) then it will restore the name of the server and the IP address, etc. which will be catastrophic if you put it on the same network as your current server.
M.
M.
.