Call Support and/or your Sales Rep.

I don't believe what you are trying to do is currently supported. Support/Professional Services may be able to help you.

Legacy Forum Name: General Discussion,
Legacy Posted By Username: CCDave

If you are looking at setting up a second, independent K server to help distribute the load (where each server hosts 5000 agents and with no crossover) then of course you can do this, call your sales rep at K to make sure the licensing for this is in the clear.

Now if you are looking for high availability (if one K server goes down the other will be up), you're SOL pal. Sorry! ......sucks doesn't it???? Many of us have brought this concern to Kaseya over the years and while we've heard promises, nothing has been done (at least nothing visible). Do call your Sales Rep, and submit a ticket to support and raise a fit about it. If enough of us complain about this, maybe K will finally do something about this.

Legacy Forum Name: General Discussion,
Legacy Posted By Username: boudj

boudj

If you are looking at setting up a second, independent K server to help distribute the load (where each server hosts 5000 agents and with no crossover) then of course you can do this, call your sales rep at K to make sure the licensing for this is in the clear.

Now if you are looking for high availability (if one K server goes down the other will be up), you're SOL pal. Sorry! ......sucks doesn't it???? Many of us have brought this concern to Kaseya over the years and while we've heard promises, nothing has been done (at least nothing visible). Do call your Sales Rep, and submit a ticket to support and raise a fit about it. If enough of us complain about this, maybe K will finally do something about this.

Isn't it possible to have a highly available kserver with vMotion? It is not a cheap solution but if you are at the point of needing an HA'd kserver then you should be able to afford the expense.

Legacy Forum Name: General Discussion,
Legacy Posted By Username: raymondm

I was going to use CA XOsoft HA (awesome product) for high availability (synch SQL and data files between both servers), but how do you deal with support at K if they do not approve? How do you deal with licensing? Also, you were looking at having a second server to also be used for distribution, once again needing K to sign off on it (licensing). Many of us are definitely good enough to come up with some redundant solutions, but we still need K involved with the process (or just plain approving of the HA process).

My last though was to have a high end SAN on the backend, and two host Virtual servers going with VMware motion in play. That way if one server has a problem you could move to the second pretty easily. But this would not solve your offloading of resources to a second server.

Legacy Forum Name: General Discussion,
Legacy Posted By Username: boudj

I am struggling with the same issue...surprise!

What I want to do, is have a standby server in another datacenter. Install Kaseya on that server and then disable all of the Kaseya services.

Then, setup transactional replication using MS SQL 2008 Standard at each end.

The theory, if the primary server fails, we redirect dns (5 minute TTLs) to the backup server, and then fire up the services.

Failing back (since I want a one way replication to avoid any corruption) would mean backing up the database on the 2ndary server and then restoring it to the primary. Reconfigure replication, and you are ready for the next disaster.

What's the problem with that method?

SQL replication (of any kind from what I know) requires primary keys in each table you want to replicate. Why is that an issue? Somebody decided it was a good idea not to have primary keys in 1/4 of the Kaseya tables.

So, I am writing to my sales rep tomorrow to see if I can get their DBA/support to sign off on me adding some primary key / auto increment columns to those tables.

Right now, I am running a daily backup (18 gb currently) and then incrementals every hour. I use robocopy to replicate to the standby server at another datacenter.

Geez...the hassle! Thanks for listening, I feel a bit better!!

Legacy Forum Name: General Discussion,
Legacy Posted By Username: rbloch

rbloch

I am struggling with the same issue...surprise!

What I want to do, is have a standby server in another datacenter. Install Kaseya on that server and then disable all of the Kaseya services.

Then, setup transactional replication using MS SQL 2008 Standard at each end.

The theory, if the primary server fails, we redirect dns (5 minute TTLs) to the backup server, and then fire up the services.

Failing back (since I want a one way replication to avoid any corruption) would mean backing up the database on the 2ndary server and then restoring it to the primary. Reconfigure replication, and you are ready for the next disaster.

What's the problem with that method?

SQL replication (of any kind from what I know) requires primary keys in each table you want to replicate. Why is that an issue? Somebody decided it was a good idea not to have primary keys in 1/4 of the Kaseya tables.

So, I am writing to my sales rep tomorrow to see if I can get their DBA/support to sign off on me adding some primary key / auto increment columns to those tables.

Right now, I am running a daily backup (18 gb currently) and then incrementals every hour. I use robocopy to replicate to the standby server at another datacenter.

Geez...the hassle! Thanks for listening, I feel a bit better!!

Why not make it simpler with VMware vMotion?

Legacy Forum Name: General Discussion,
Legacy Posted By Username: raymondm

raymondm

Why not make it simpler with VMware vMotion?

We need to keep the Kaseya system separate from our VS farm and SAN. Right now, we can sustain an entire failure of our vs farm and san and still have Kaseya live. Alerts are sent through multiple smtp servers outside of the cluster as well.

We have another system that monitors our Kaseya server to make sure it is online and responding to requests properly on both the kaseya tcp port and the web service.

The goal is to know if a major disaster happens.

The other issue is cost, setting up a new cluster of servers/san is a bit costly.

Legacy Forum Name: General Discussion,
Legacy Posted By Username: rbloch