Home Forums Discussions Problems V5 has Critical Bug - frequent drop at SQL server

Hi Bert,

I truly appreciate your help and advises. After all, spending time to answer IT questions and helping the other physicians after a whole busy day of seeing patients is a noble act. It was a long day yesterday. I was so frustrated last night when I log in after midnight. All of a sudden, everyone started to say that AC is built for small physician groups using a few computers only, as if connecting 30 computers to the server is unbelievable. I suppose pushing the limit is healthy for long term growth of AC.

To answer you questions, there are no specific laptops/towers that are more likely to have the SQL drop. Ordering tests and closing and reopen AC will put any client at risk. When the system is unstable, its kind of like you have to wait at the end of the line for your turn to get back in the system again. Rebooting the server did help.

Knowing that other groups with higher number of computers are having the same problem points to the problem of AC V5 and SQL, but not our network issues. Your suggestion of trying the SQL workgroup edition and contacting the SQL expert will be my next step. However, both are beyond my personal capability. I have to ask my IT guy to do them. I am not sure AC IT will guide us through the process of switching to the workgroup edition. Can you outline the procedure for us?

Hi Philip,

I completely understand your frustration. We had a similar problem years ago with two doctors, and you have more and in different locations.

I definitely will help. I will outline the procedure for you. I will PM to you tonight. It would be easier to email if you would be so kind as to enter it in the PM. You can disguise it so bots can't harvest it, by doing something like badams[AT]riverviewpediatrics[DOT]org_remove or anything like that.

Try to enjoy your Friday.

After cleaning out our one overfilled mailbox, I got shut down only once today.....as opposed to every patient on wednesday.

Philip,

Still plan on working with you on this problem. However, I just want to make sure it isn't this issue with the messages. StLawrences' message sounds rather promising, although I still don't understand why it shut down at all.

One thing that would be helpful from you and StL and everyone else, would be are we talking the same thing here or apples or oranges. In Philip's case I am hearing it is a drop from the SQL Server. Some are saying the program "shut down." My program crashes every once in awhile, but I think of it as more of a local issue than a database issue.

I am not well versed in computers. I was suddenly having problems with Amazing Charts closing as I was with a patient usually in the process of sending orders and I would have to restart Amazing Charts and fortunately the patients encounter wasn't lost although I would sometimes have to redo my orders. It was to the point this week that I was going to call AC or email them as it was becoming difficult to get through my days work....getting behind with office hours,etc. My husband who is the self taught geeky computer guy in my office came here and found Jon's comment and we looked into it. As I mentioned I was sending all my orders to labdoctor's mailbox. Looking at our work processes there was no need to do this and we just had a very full mailbox that no one was using. Today was the first day of office hours since we went through and deleted all those useless orders stored in an unused mailbox. There is a tracking system in V5 that makes use of "labdoctor" redundant so we are not sending those orders anymore. Just saving. I'm not sure if Philips problem is the same as mine was but just adding my $.02 just in case.

Well, I am not a SQL Expert, and this would probably be easier for me, if I had v5 in production. Similarly, the time my computer crashes the most is when I am doing a letter and doing the Print Instructions (Plan). But, it isn't losing connection to SQL Server. I would think that if one were actually losing connection to their databases, AC would not only crash but when you restarted, it would bring you to the window where your browse to the database.

Just talked with my husband. We do not disconnect from our database. The program just shuts down and has to be restarted. When restarted it will bring back the note I was working on. This week it was happening with every single encounter. Happened only once today. I am using V5 because I get really busy in the summer and I don't want to wait until then to upgrade to the new version. V5 is alot different and takes some getting used to and I can't deal with learning it when things get busier.

Count us in on the disconnecting SQL. My dad's office is setup very basic for a Microsoft Windows based network. Server is a home-built system with an Intel Core 2 Duo 2.13Ghz, 8GB DDR2 RAM, Four Seagate 10k-rpm 136GB SAS disks on a 3Ware 9690SA-4I running in RAID-10 with write cache enabled. Windows Server 2008 Standard 64-bit is installed. The entire machine is built for speed. We have 7 other computers with AC installed to connect to that main server. The other computers are a mix of wired desktops and wireless laptops. We are using Symantec Endpoint for virus and intrusion protection.

We continually have users get kicked out of Amazing Charts and presented with the "browse to database" screen when they try to get back in. Running Amazing Utilities and repairing connections seems to do the trick and then everyone can get back in. My dad is one of the culprits who has about 300 messages in AC inbox, so perhaps it is the same issue as the slow V5 thread. I'll know soon hopefully as my dad is cleaning out his and the other staff inboxes. After that we will try to get brave and upgrade to a 180-day trial of SQL Server 2008.

Is there any reason why AC isn't web-based yet? I would think going to a web-based platform would not only allow you to keep things very stream-lined (think very thin clients) but you wouldn't have the issues of AC disconnecting as much from SQL because that is all handled server-side.

I would LOVE to see AC support MySQL, as that would allow almost everyone to use a very rock-solid and performance driven SQL server without shelling hundreds (or thousands!) of dollars to Microsoft. Personally I will be looking into changing all of our desktops to Linux machines and using RDP to connect our users to this single server for access to AC. I'm just tired of the issues in the "peer to peer" and the "client to server" AC environment.

What version of Version 5 are you running ? It seems most people who are on 5.028 have been having good results. When you say continually - how often - 10 times per day ?

In regards to the web based - in a lot of our cases this would not be ideal as internet goes down, etc.

Adam,

Hi. Thanks for posting. It is obvious from your post that you know what you are doing, maybe more so than anyone on here. Therefore, it is a bit surprising that you come to the conclusions that you do. While SQL 2005 Express isn't SQL 2008 or Oracle, it is more than enough to handle seven clients and a few others. I may also be completely mistaken (probably am), but I am not sure how compatible it is with 2008.

I think there are always a few of us who for whatever reason have issues in the beginning with AC and SQL. It's usually connectivity. I don't know why you are having these issues. I am sure firewalls aren't an issue, although DNS and other things could be a problem.

I can see that your setup is way, way more than anyone would ever need with AC. From reading the board, you will see that there is no one who likes Client/Server better than I, but truth be known, it will run on Windows 98 on a 512Mb peer to peer. Maybe not blazing speed, and I wouldn't run the database on that, but you get what I mean.

In answer to web-based, it will never go web-based. Jon has made that pretty clear. There are ASP web-based models and there are local application models. One doesn't usually migrate to the other. Usually. As far as MySQL, I don't think Jon is moving in that direction either. Some people talk about web-based as if everyone is having these issues. I think it is the small minority. Of course, if you are one of those, it is tough. This may be a dumb question, but have you talked with support? I know they can't always fix the question, but they will stay with you until they do or at least try to.

As a user from the get go and Access, I have been lucky with the SQL version, but it has been a growing pain for some. I don't have the time to look now, but there was one group having issues that recently changed to SQL 2005, but I don't know the outcome.

Do you have any IT people there or someone who specializes in SQL or a DBA?

As for web-based I was meaning that the local server would serve the actual "web page". So instead of installing AC on each network computer you would just browse to the host/IP of your main server and access it that way.

I'm the "outsourced" IT guy actually. I know more about networks and servers than I do about medical stuff, that's what my dad is for I've done some pretty vast testing. At the time the nurses and scheduling folks get booted from AC I always have them try to browse a fresh web site. I will often times just have them go somewhere random, like viennabeef.com and so forth. The webpages always load up just fine and fast. At the same time I will RDP into the server and double check DNS requests, pings and traceroutes out to various sites on the Internet as well as to computers on the network. Everything checks out. The network computers get the "browse for database" screen when re-starting AC. They can browse to the network share, click the xml file and AC accepts it then tries to connect to the server, just to time out again. The only things that fix the issue is rebooting the server, restarting the SQL service or going into Amazing Utilities and hitting "Repair connections". After one of those actions is done then people are able to connect again.

Unfortunately I am a CCNA, not a DBA. My experience with SQL Server is fairly limited so it's a new animal to me. The tough thing is that we are stuck with Server 2008 standard, I can't downgrade to 2003 Standard due to funding. In some other threads you were toying around with upgrading 2005 Express to 2008 Express, have you beat that around any more? Does it seem to survive ok? I may give an upgrade to 2008 Express a shot.

I am currently having staff keep record of the date/time, error message if any and what they were doing at the time of an AC "crash" when it presents them with the "browse to database" screen. Hopefully I can compile that data at the end of this week to get a better idea of how often this is really happening. The Event Logs don't show anything so far that I've noticed that would indicate an issue with SQL Express, but who knows. We are using AC 5.0.28 currently.

Check out these 3 scenarios: http://blogs.msdn.com/b/sql_protocols/archive/2005/12/19/505372.aspx

I agree with getting a SQL specialist- it may be a pretty easy thing for him to diagnose/fix, but not for us.

Al

I have found that keeping Amazing Utilities "live" on the main computer, and hitting the "Repair Connections" when clients can't find the database is the fastest.

I also find it interesting that at the same time that client computers are seeing the error message, other, already logged in clients are accessing the database without interruption. However, if a client logs out and then back in -- error box. This must mean something to a SQL expert.

Adam,

Again, sorry for your issues. I am not sure if you will be able to simply upgrade to 2008 Express or not. It all depends on the program. AC is designed to install and make a default instance to SQL. I have tried running a different SQL and not installing SQL from AC and making an instance. It didn't seem to work. Of course, I am not guru either. Not even close.

I guess my point is whenever these issues crop up, many on here just assume it is AC or SQL. Granted, the problem is you and they are losing connectivity, but there are so many variables.

It is difficult, I am sure, for AC support to know all of the different variations of networking out there.

I have two SQL gurus I use who kindly help with free emails, but charge me around $75.00 per hour otherwise. They are very good. They did try on here, but I am not sure where that went. I just know I got in the middle too much and muddled things up.

Probably the best way to go would be to go to experts-exchange.com and post your question. Put it under the SQL section and the networking section and see if they can help. If you go that route, I will steer my two friends to your question. You get six trial days. I pay for mine. Given it is a trial you can't offer as many points (I know, crazy, but that's computer geeks for you), so you may get less bites, but I can get my friends there for very few points.

The thing about Experts Exchange is you can ask a question and get four experts who hang in there with you for three days, or you can get one who gives up after two posts. But, you are guaranteed to get a better and quicker response there than on most any other computer help boards, and it is set up far better than any other. If you got a couple of experts to try to solve the problem with you, there is a good chance they may be able to.

I am hoping that, now that the 364 error is solved, that this SQL database access error will be a focus for the next update. It happens to many of us with all sorts of networks -- just read the 6 pages of posts already on this topic!

OK, here is a complete guess. And, I am not sure why this is.

But, in the system databases Master, Model and MSDB; autoclose is set to false. In the user databases, all have autoclose set to true.

Now, I am not sure if this would make a huge difference in any office with a small amount of users, but I think most databases are set to have autoclose false by default.

If autoclose is true, then when the last person disconnects from the database, SQL closes the database and releases all of its resources back to the server. When someone reconnects, the database opens and the resources are recommitted. This all sounds well and good and probably is.

But, if the next user is a split second after the first user disconnects, then SQL must make an about face and retrieve the resources. This can and does slow down the database. Whether or not it could make it crash, I have no idea.

Hello,
I've been following this thread since its inception as the network I support is very similar to the original one described in Philip's post. I've advised not to upgrade yet because it seems as if the issue affecting those who have joined in this discussion hasn't been fully identified and is not yet fixed. Has Philip or anyone else who has chimed in successfully found a fix or received any promising info regarding the network disconnections after the upgrade?
Thanks!
Tom

I think it may be better since version 5.028. In my network -- Win Server 2008/R2 server using workgroup config (not domain) -- I have had fewer total disconnects (down to once a day). Still plenty of "freezes" when trying to access the database, such as opening a patient's chart or an Imported Item, lasting a few seconds or more (the blue circle on Windows 7).

I know this post is really old but we are having this same problem in our office. We are getting 4 to 5 drops a day. We have a total of about 15 computers accessing a very strong server that we just purchased. We had this problem in the past and figured a new server would fix it and it didn't. We currently have a full SQL system not express so that cant be the problem. I have been doing a little research on this problem and have found that the server does a auto disconnect after a period of time. I have changed the setting in the server to turn off auto disconnect. Has anyone else tried this and does it work? I am crossing my fingers.

Hi salford,

Still need a lot of information. What is the serve? How much memory does it have? What are your clients? How were you able to use SQL Server 2005 and not Express? Big question: Wired or wireless?

The statement that you have a full SQL system so that can't be the problem is not necessarily true. It could still be SQL.

I've had the same problem since I went on 5.17 a few weeks ago. have 16 clients and a win 2003 server. No problems with 5.028 but dropped connection to the SQL server on several stations a day. very difficult to reconnect at times. How do you turn off auto-disconnect?

Are you using mapped drives by any chance? Jusst a thought.

The famous Bert replied to one of my post. I read up on these boards all the time and see you Bert I feel honored that you replied to my post. :-) I appreciate it. Now on to answering question. Our IT guy set it up where we are not using express but the complete SQL. I ask him to do that from reading your post Bert. That is what I mean. Our new server has 8gb ram 3 X 250 memory Dell power edge running server 2011. Everything in the office is wired, All of our clients are running xp. There is a total in office of close to 15 and we have 2 people that work from home who remote in. Now mapped drives that question I do not know the answer to. Please let me know if they need to be mapped or not. I will call the tech guy and get him here if that is going to fix the problem.
The (auto disconnect) I did yesterday on the server; not sure if it worked but I didn't hear anyone in the office yell out "ughhhh restart again" but I will see today.
You turn off autodisconnect by going in the server to the command prompt and type NET CONFIG SERVER/AUTODISCONNECT:-1 , When we had a drop I did check out my log on the server yesterday and it said MSSQL cant except a new connection because it is shutting down. No one here shut it down. Maybe this is the cause of it. How do I fix that?

ok IT guy just told me mapping is done via login script whatever that means

Thanks for all the kudos. I think you have actually hit on something here. I will email my networking guy and see what he says. All of the initial research I did talked about auto-disconnect with Mapped Drives, hence the question. AC doesn't work well using Mapped Drives, if at all, so I wouldn't recommend it.

After further research, I found that auto-disconnect (which seems to server no purpose other than to disconnect users who do not log out to make backing up databases safer) can disconnect LAN connections.

As you state, after a certain amount of time based on the -1 up to 65,000 or so, a LAN connection will disconnect. Supposedly, trying to use it again will reconnect it.

While you can change the parameters as you have (actually turning it off) via your command prompt which is recommended, it is actually better, according to my reading, to use the Windows Registry directly.

HKLM\\System\CurrentControlSet\Services\LanmanServer\Parameters

find Autodisconnect and change the value to something like 0xfff. This is generally used for a domain.

I am rather impressed that you found this. It may solve a lot of people's issues.

The only thing I would change since SQL 2005 can use as much memory as your server will take, would be to up it to at least 12 or more to increase your buffer.

While you can do that, I am just talking about going to a window such as My Computer or Windows Explore, choosing Map Drive at the top and browse to a folder on the server. It is then seen by your computer as its own drive. It will reconnect on login if set that way, which is the default. There are many reasons to map a drive. Working with AC isn't one of them.

I happen to chat with Jesse from AC (for another problem) and the pc he was working on did the sudden disconnect. He then looked around and initially said my netframe is old (it was 2.0)so we upgraded to 3.5. but when I upgraded another pc with the 3.5 it also did the same thing. So this wasn't the issue.
What Jesse did was chose the Ac xml file using the static IP address of the server it was on and this worked. I will let you know if at least this solution will stick.
I will also try turning off the autodisconnect.
Thanks.

Rene
IM Oswego, NY

Well yesterday was the first full day with the autodisconnect crap on -1. "Wait I have to wipe a tear" It didnt work for us. ughhhhh. I am going to try the reg change Bert found and see if that happens. Also Today I am paying our IT guy to work with ac tech support to see if they can figure it out. My tech guy said that SQL error about auto shut down could be the issue.


I hope to solve this problem before the conference next week so I can sell the fix to everyone when I am there. :-)(Joking)


Let me know how you turn out with your experiment Northcountrydoc

Hi sal,

It will be more interesting to see what the value is on the registry. If it is set to not autodisconnect.

Please see my private message.

The workstations have stopped seizing for the past few days. I don't know what did it. I did a few things: upgraded all clients to netframe 3.5, used the IP address # instead of name to point to tne main AC database location (I use static IP addresses in my network) and Jesse from Ac said this may help, changed some of the older pcs to faster ones and increased their memory to minimum of 2 gig. I did not change the registry autodisconnect parameter. Hope this helps

Rene

I still don't understand the IP address and static IPs. You still have to connect using the .xml file.

We had a lot of similar problems last week. Then the server started crashing regularly. We changed out the power supply and it was much better today, but we were half-staffed so there may be problems in the future.