Home Forums Discussions Problems Automated processes are failing

We had someone from AC support upgrade our SQL Server from 32-bit to 64-bit on Friday afternoon.

Since then we've been receiving these "An automated process failed" messages every few minutes. Related to UpdoxSecMsgTask, and MedFxOutSvcTak. Looks like just those two over and over.

I tried rebooting the server, but the messages continue.

I just wrote AC Support but thought I would try the forums also since it is the weekend.

Thanks for any ideas.

Wow. Big step. Which SQL Servers are you using? Were you having performance issues with 32-bit? Moving to 64-bit SQL can be beneficial, but it can also give no performance increase and cause problems. Is your processor fast enough to handle 64-bit SQL.

There can be some issues with some 32-bit programs working with 64-bit SQL. Especially issues with 32-bit drivers. I am surprised that AC support did this. I will be more surprised if they can fix it.

My recommendation would be email my friend who is Microsoft certified in ALL versions and all architectures of SQL. Many times, it is necessary to map out the exact plan of changing 32-bit instances to 64-bit. I will forward your question to him and see what he says.

I'm inferring from this that you have both of those interfaces active?

I would start with going into the admin side of AC, inactivate the interfaces, restarting the AC services, see if the error stops, then activating them again. Next I would try uninstalling the interfaces, then re-installing them.

Curious - how many users you have on your network connecting to the AC server?

We have 3 providers and 7 total users. We were getting abysmal performance at times, as well as timeout errors. AC Support said he thought this was related to 32-bit SQL Server.

We're not sure yet if it has solved the problem. Get some feedback this week as people use it for the first time.

I don't see how to de-activate a service. Clicking on Interfaces I see a list of "current interfaces":
MUIR is Active
MUPHR is Active
NEWCROP is Inactive
Piedmont Medical Laborate is Inactive
UpDox is Inactive
X-Link HL7 is Active

My errors are from UpdoxSecMsgTask and MedFxOutSvcTak.

So seems like UpDox is already Inactive, I don't know how to activate it. Double clicking on Updox in the list on the left just shows a splash screen for it. I don't see activate/de-active controls anywhere.

And MedFxOutSvcTak I'm not sure what that corresponds to.

I guess at this point I will see what support says in the morning. These errors started when they touched the system, so it's something they did. It seems like AC is still running overall so it's not a crisis.

$10.00 says AC support will say they didn't cause it. This should have been fixed over the weekend. Not sure what it will do to MedFx, but that is a billing company.

From what you said, it sounds like AC support just guessed it was due to 32-bit SQL. That makes little sense to me. There was no other attempt to look at the network first? Like is there a 10 or 100 Mb/s switch? Why was SQL looked at as the culprit. We have more users than you, and we were still using 32-bit 2005 Express, and it flies. I can't imagine SQL causing performance issues in too many offices.

Here is the deal. I am not sure what the semantics are, but you cannot upgrade from 32-bit SQL to 64-bit SQL. You simply can't. The 64-bit drivers just won't be there. Your fist move would be to try to find 64-bit drivers for the two processes. The only way to do this properly is to uninstall the 32-bit SQL and install the 64-bit SQL. Is this what AC did?

AC makes a product which will run on Windows Server 2012 Standard R2. That doesn't mean they know that OS. They make a product which uses SQL Server for its database. They know what tables houses certain data. But, they don't do support for SQL Server.

A task like that should have been done by someone who knows SQL Server. Someone who can tell you if you will get a performance boost in the first place. I suspect you will continue to have error messages without the proper support fix.

See PM for further detail.

Must be fixed.