Hi all,
I’m relatively new to Epicor ERP 10 and recently just went live running on 10.2.400.11. I currently have 1 App Server and 1 DB Server respectively.
Has anyone created a secondary app server to be a backup server just in case the primary app server is down? I’d like your thoughts and advice on it before I commit to a secondary app server. My ERP 10 is hosted by a third-party cloud provider for your information.
We don’t have a 2nd app or database server on standby specific to Epicor. With our backup solution we can spin up a new app server VM and restore the database in just a couple minutes. So we don’t have a need to have dedicated standby servers. In the event we couldn’t spin up new VMs we could run the app server and database right off our backup appliance.
We have two app servers as well as the DB server, but not for redundancy. We just have two running in parallel because our users tend to be very demanding about printing and submitted processes happening quickly, and having two provides a sort of continuous redundancy.
So as far as that goes I can confirm that having more than one app server works, but can’t offer much advice on whether it’s a good solution as a backup. In our case it wouldn’t be, because the clients all point to the main app server anyway, but it would depend how you set it up.
We also have two app servers for the same reason. One is dedicated to running the task agents. The other is the main app server that clients connect to. And where EWA runs.
When you all are referring to “2 App Servers”, we’re talking about independent boxes (or VM’s). And not just multiple App instances on the same box, right?
If you had App PRD1 running on \\App01, and App PRD2 on \\App02, doesn’t the client install point to a specific App?
So if you had users using PRD1 (on \\App01), and it went down, wouldn’t you have to redeploy the clients (or at a minimum tweak their configsys files) for the users to connect via PRD2 (on \\App02)?
Or is there some way to have automatic fail over?
So what happens in this case if the app server running the task agents fail? Will there be any failover to the secondary server? What I’m hoping to look for is a solution where if my primary app server fails/hangs, my operation can be still be running as per normal as it will be falling over to the secondary app server.
I’m actually talking about 2 independent VMs so if App PRD1 fails, App PRD2 will automatically take over. Basically it’s just like HA.
Are you talking about a complete server failure or just the task agent stops running?
It’s easy to have auto fail over for VMs. But I’m not sure how you would handle the client. The hostname of the server wouldn’t be the same most likely. So the client wouldn’t know where to connect. You could use a CNAME for the hostname and change that in DNS I suppose. But not sure how that would work with Epicor. I’m thinking you would need to modify the config files to point to the CNAME initially.
What kind of infrastructure is used by your cloud provider? Have you ask them what kind of capabilities they perhaps already have in place. Several times if your provider is basically just a go-between for Azure, AWS, or on a VMWare platform they can offer some affordable options, if not free, for protecting you in case of a app server or SQL failure. If they don’t have a good economical solution I would question continuing to host with them.