I am using a Server class machine with Windows 2000 Server. This machine is home for my lookout processes. I have a server app (unlimited I/O AND unlimited client). The client app (using all symbolic links) is also running here as well as on multiple client nodes. This server machine is also serving the web client app. This machine also is used for various other programs including MAX, internet explorer etc. Occasionally when I am using another program, Lookout just quits serving data to the client nodes. I will look over at another monitor and the displayed data will not be changing. I will maximize Lookout on the server machine and the data is still changing.
I will then shutdown Lookout on the server, of course red Xs appear t hen. I will restart Lookout and the system will recover. If this happens again (Lookout quits serving) I will try the same remedy. However, when I restart Lookout this time, the system does not recover. The red Xs on the clients remain even after the server is running again and on top of that, Lookout on the server will quit logging historical data, meaning all my trends are now blank. In order to restore the system, all Lookout clients must be closed and the server machine must be re-booted.
On another note.... I have 3 client nodes here in front of me. Two of them, I have checked the option to run as NT service so it will start automatically. The 3rd, will not accept this option. I will check the box and hit OK. If I try to go to Options>NTService after that, Lookout generates errors and shuts down. If I restart Lookout, the NtService option is again unchecked.
OK, it appears that my server process just quits saving the state file periodically. It does not seem to have anything to do with any other programs. Since there is nothing at all on this site about the state files, I need some help ASAP. I have the state file saving every 10 minutes now, so I can better tell when the problem occurs. The last time I fixed this problem (by rebooting the server) it took 3 days for the problem to re-occur. That happened this morning, since then it has happened 3 more times within the last 3 hours. Also, a couple times this morning, my client nodes came up with an alarm saying "error saving the state file", this shouldn't happen on a free client should it? That alarm NEVER appears on my server.
New twist, this morning my app is once again not serving data to the clients but the state file is still saving as it should. Also, all of my client machines are now showing the "error saving state file" alarm.
I get the strange feeling I am talking to myself here.
You wouldn't be running your server file on a computer with 2 hard drives, would you?
Reason I ask is I just got 2 new Dell 650 workstations, each with 2 hard drives, and I am having problems similar to yours.
I installed Lookout 5.1 on the 😧 drive and everything works great from the server to the plc's, but from the server to the clients I am having problems similar to yours.
First, I noticed that my trends were dissappearing, then after a while communication would just stop without any red "x's" coming up. It's like the clients see the server but no data can be transferred between them.
I use a symbolic link to connect my signals from my client machines to my server and when the problem occurs I can access my symbolic link at first even though no data is being transfered, then I lose my symbolic link, but if I go up to "Network" in object explorer I can hit the server for a little bit longer before I lose that connection "route" - expalin that one.
Also, since I installed Lookout on the 😧 drive I wanted the database on the 😧 drive - well, Lookout decided it would also create a database on the C: drive which is what I thought was causing my problems. But, after disabling one of the drives the problem still returned.
Also, I noticed earlier today that my state file stopped being saved as well so I really don't know what is causing the problem. It's driving me nutz!
I called technical support and they are setting up a test center on a computer with 2 hard drives to see if they can duplicate the problems I am having.
Glad to hear from someone else. I am using a Dell PowerEdge 2600 Server class machine with Windows 2000 Server O.S. I have found out from NI that Lookout has not been tested on Server class O.S. I do have 2 hard drives, but I am running everything on C:. I have a 10/100/1000 ethernet card in a 64bit PCI slot for my network connection. I am also running an Allen Bradley dual channel DH+/DH485 card. I have PLCs on the ethernet network, DH+, DH485 and also 1 on Serial port.
One thing I did was to copy my process files to another sub-directory then delete the originals. I ran the new apps for a day before returning them to the Lookout5.1 root directory. (This is where the apps must reside in order to use webclient). Since then my st ate files have been updating regularly, but I still keep losing data to the clients. Right now I am running a second server process with a single waveform object. I creadted another client app with a sym. link to the waveform and connected it to a pot. The next time it fails, I should be able to tell if it is process specific (if only 1 loses data) or if it's something with the computer and/or NI software services.
Jason, I share your frustration, but it's nice to know I'm not alone in the world with this!
Have you noticed a database being created on the second hard drive as well as on the C: drive?
I went back to my older machine today (Dell 340 workstation also running Windows XP) which has only one hard drive and I am having no problems which leads me to believe the computer is causing my problem. I plan to uninstall the second hard drive on Monday and give it another try. If it still fails, I am going to turn my attention to the network card (swap them out - the 340 has a 3com card and the 650 has a Intel Pro Set 1000 card).
Since both of these computers are running the same version of Windows XP I cannot see that as a cause of the problem.
We'll get to the bottom of this eventually, I hope.
Yes, I am sharing alarms between the server and all clients. I also have event objects within my client app.
This morning, my clients were once again not receiving data. I have a second app (server) running a waveform and am reading that value on a client. That app failed as well. So this should indicate the problem is not my process.
My 2 hard drives are slightly different. Since it is a server class machine, the 2nd drive is a mirror backup. The main drive shouldn't know it's there. Basically, everything that exists on my main drive is also on the backup drive.