I am running HM as a service (but I am still logged on to the console of the virtual machine). Every few days the program crashes. The error message is:
Runtime Error!
Program: C:\Programme\HostMonitor8\hostmon.exe
This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Even more of a problem is that while this message box is on the screen, the watchdogs still find the host monitor task and are perfectly happy with it.
System: Windows XP Pro, Version 2002, Service Pack 3
Host Monitor: Version 8.26
running in a VMWare Virtual Machine: Version 4
Console opened in VMWare Infrastructure Client: Version 2.5.0 Build 104249
My desktop computer: Windows Vista Business 64bit Service Pack 2
Host Monitor Service Crash
This means HostMonitor is not crashed. It does respond to requests and send data to WatchDog so HostMonitor works fine.Even more of a problem is that while this message box is on the screen, the watchdogs still find the host monitor task and are perfectly happy with it.
I don't think HostMonitor shows this error. It looks like this error is recorded (or displayed?) by some 3rd party software. Probably you are using Shell Script test method that starts some 3rd party application? If this application does not finish its job within specified timeout, HostMonitor will kill that process.Runtime Error!
Program: C:\Programme\HostMonitor8\hostmon.exe
This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Regards
Alex
Sorry, I forgot to mention that when I click OK on this dialog box, the watchdog starts barking and the HostMonitor service is stopped. I have to start the HostMonitor service manually again. This led me to thinking that the watchdog just looks for the existence of the HM process.KS-Soft wrote: This means HostMonitor is not crashed. It does respond to requests and send data to WatchDog so HostMonitor works fine.
I don't think HostMonitor shows this error. It looks like this error is recorded (or displayed?) by some 3rd party software. Probably you are using Shell Script test method that starts some 3rd party application? If this application does not finish its job within specified timeout, HostMonitor will kill that process.
I do actually have one Shell Script test. It tests a certain value inside a file. It is dependend of a File existence test. It has been working fine so far, giving proper results in both cases, good and bad. I don't think the script is to blame.
There seems to be a Microsoft knowledge base article (http://support.microsoft.com/kb/884538) about this problem, although it applies only to XP SP 2 and I have SP 3. Do you think I should try to install the hotfix anyway?
Is there anything else I can do to fix this?
No. WatchDog sends request to HostMonitor on regular basis. If HostMonitor does not respond, WatchDog starts alerts.Sorry, I forgot to mention that when I click OK on this dialog box, the watchdog starts barking and the HostMonitor service is stopped. I have to start the HostMonitor service manually again. This led me to thinking that the watchdog just looks for the existence of the HM process.
May we see that dialog box?
Not sure. Could you try to disable this test item for testing?There seems to be a Microsoft knowledge base article (http://support.microsoft.com/kb/884538) about this problem, although it applies only to XP SP 2 and I have SP 3. Do you think I should try to install the hotfix anyway?
Do you use some other functions (like "External" test or "Execute external program" action) that launch 3rd party applications created by Visual C++?
Regards
Alex
Here is a screenshot with the dialog box:
I found a recommendation for this error to reinstall the dotnet framework, which I have done meanwhile. I also changed the shell script test to make sure it doesn't crash when the file is not available (I hadn't done this before because the test runs dependant on a file existency test).
As the problem occurs at irregular intervals, I will observe this and wait for it to happen again. Next time it happens I will try to find out which test ran last, before restarting the service.
Best regards and thanks for your support,
Juergen
I don't really want to do this. I need this test once every morning.Could you try to disable this test item for testing?
Nope, sorry.Do you use some other functions (like "External" test or "Execute external program" action) that launch 3rd party applications created by Visual C++?
I found a recommendation for this error to reinstall the dotnet framework, which I have done meanwhile. I also changed the shell script test to make sure it doesn't crash when the file is not available (I hadn't done this before because the test runs dependant on a file existency test).
As the problem occurs at irregular intervals, I will observe this and wait for it to happen again. Next time it happens I will try to find out which test ran last, before restarting the service.
Best regards and thanks for your support,
Juergen
Message looks like its HostMonitor fault. But HostMonitor does not use Microsoft Visual C++ Runtime Library!! So it must be some 3rd party software that uses C++ Runtime Library.
How it can crash HostMonitor? Not sure...
Probably its not Shell Script test. May be its ODBC or Active Script test. ODBC drivers work in address space of the host process. So any bug in ODBC driver can crash HostMonitor.
Do you use ODBC logging or ODBC test method? If yes, what ODBC driver do you use?
Active Shell tests?
Do you have installed some antivirus monitors, personal firewall, content monitoring software? Non stanard winsock components?
If you do not want to disable that test, please send e-mail to support@ks-soft.net, provide your registration name and/or order number. We will provide "debugging" version of HostMonitor, it may provide a little more information regarding the problem.
Could you disable ODBC logging and tests (if you have one) and install HostMonitor on "clean" system without antiviruses?
Regards
Alex
How it can crash HostMonitor? Not sure...
Probably its not Shell Script test. May be its ODBC or Active Script test. ODBC drivers work in address space of the host process. So any bug in ODBC driver can crash HostMonitor.
Do you use ODBC logging or ODBC test method? If yes, what ODBC driver do you use?
Active Shell tests?
Do you have installed some antivirus monitors, personal firewall, content monitoring software? Non stanard winsock components?
If you do not want to disable that test, please send e-mail to support@ks-soft.net, provide your registration name and/or order number. We will provide "debugging" version of HostMonitor, it may provide a little more information regarding the problem.
Could you disable ODBC logging and tests (if you have one) and install HostMonitor on "clean" system without antiviruses?
Regards
Alex
Last night I got the error again. This time I checked things before restarting them...
I have a (refreshing) browser window open for host monitor webservice:
The host monitor shows this (note that the systray icon is still moving):
The last entries in host monitor log file look like this:
It seems you may be right about the ODBC check. There is another ODBC check that has last been logged at 00:00:20 and had the result OK.
The last entries in host monitor syslog file look like this:
Note the IP address is my desktop computer - the one that has the browser open and refreshing. The last connection was at 03:28:05, roughly 30 minutes after the last logged ODBC test. Both ODBC tests run every 10 minutes. The log is written only when the status changes (Midnight setting). I changed this option now to "Full", maybe we get more info.
The ODBC driver in question is the driver supplied with the Oracle instant client, Version 10.02.00.04, file SQORA32.DLL from 03/03/2008.
I am not aware that I am running any Active Shell tests.
I will do as you suggested and ask the support for a debugging version of HostMonitor. I wouldn't like to disable the ODBC tests, they provide valuable information. The problem is that the tests are running every 10 minutes and everything seems to be fine for days. Because of the irregularity of the error, even if HostMonitor ran fine without the ODBC tests for a few days, we wouldn't really be sure if the reason is the ODBC driver.
Best Regards,
Juergen
I have a (refreshing) browser window open for host monitor webservice:
The host monitor shows this (note that the systray icon is still moving):
The last entries in host monitor log file look like this:
It seems you may be right about the ODBC check. There is another ODBC check that has last been logged at 00:00:20 and had the result OK.
The last entries in host monitor syslog file look like this:
Note the IP address is my desktop computer - the one that has the browser open and refreshing. The last connection was at 03:28:05, roughly 30 minutes after the last logged ODBC test. Both ODBC tests run every 10 minutes. The log is written only when the status changes (Midnight setting). I changed this option now to "Full", maybe we get more info.
The ODBC driver in question is the driver supplied with the Oracle instant client, Version 10.02.00.04, file SQORA32.DLL from 03/03/2008.
I am not aware that I am running any Active Shell tests.
No, this is a standard WinXP SP3 with nothing else running on it. Windows Firewall ist active, though. Firewall exceptions are Advanced Host Monitor, Network diagnosis for Windows XP, Remote desktop, Remote support, rma, and Web Service for HostMonitor.Do you have installed some antivirus monitors, personal firewall, content monitoring software? Non stanard winsock components?
I will do as you suggested and ask the support for a debugging version of HostMonitor. I wouldn't like to disable the ODBC tests, they provide valuable information. The problem is that the tests are running every 10 minutes and everything seems to be fine for days. Because of the irregularity of the error, even if HostMonitor ran fine without the ODBC tests for a few days, we wouldn't really be sure if the reason is the ODBC driver.
Best Regards,
Juergen
We sent you a link to test version of HostMonitor. Please install official release 8.28 before using that update.
However I doubt it can help
This problem is caused by some 3rd party software (DLL) written in Microsoft Visual C++. If you will not disable some tests/options one by one, how you can find what exactly software leads to the problem?
Google did not find anything similar in relation to Oracle ODBC driver.
BTW: Do you use "Send message to pager" or "Send message to beeper" actions? These actions use some 3rd party DLLs. Though we never heared about such problem related to these actions.
Regards
Alex
However I doubt it can help
This problem is caused by some 3rd party software (DLL) written in Microsoft Visual C++. If you will not disable some tests/options one by one, how you can find what exactly software leads to the problem?
Google did not find anything similar in relation to Oracle ODBC driver.
BTW: Do you use "Send message to pager" or "Send message to beeper" actions? These actions use some 3rd party DLLs. Though we never heared about such problem related to these actions.
Regards
Alex
Thanks, I will do this.We sent you a link to test version of HostMonitor. Please install official release 8.28 before using that update.
My problem is, I have 105 tests in 53 folders, and at the moment I don't have a clue were to start. The error occurs in irregular intervals - last night everything went fine. If I disable and enable tests one by one and see if the host monitor runs fine for let's say three days, it might take months to find the culprit. I am actually hoping that, by determining when exactly the crash happens, I might be able to find the offending test.If you will not disable some tests/options one by one, how you can find what exactly software leads to the problem?
No, I don't use this.Do you use "Send message to pager" or "Send message to beeper" actions?
I will install the newest software and the debugging hostmon.exe. Also, I have now the log settings on "Full", so I hope to see the last successful test that was executed. From there I'm planning to find out what the next test would have to be - this should be the one causing the crash.
I will get back to this thread when I have more info.
Thanks a lot for your support.
Regards,
Juergen
Well, you don't need to disable all tests one by one.My problem is, I have 105 tests in 53 folders, and at the moment I don't have a clue were to start. The error occurs in irregular intervals - last night everything went fine. If I disable and enable tests one by one and see if the host monitor runs fine for let's say three days, it might take months to find the culprit. I am actually hoping that, by determining when exactly the crash happens, I might be able to find the offending test.
If you do not have some non-standard winsock/antivirus/firewall software, then you may skip the following tests
Ping TCP Trace SMTP POP3 IMAP
DNS LDAP RAS UDP NTP/SNTP
Radius HTTP URL
Drive free space
Folder/file size
Folder/file availability
Compare files
File Integrity
Count Files
Text Log
UNC Process Service SNMP
CPU usage Temp. Monitor Traffic Monitor
SNMP Trap Mail Relay DICOM
DHCP SSH HM Monitor
Also you may copy HostMonitor folder, start 2nd instance from new folder, disable logging and alerts and use this 2nd instance for testing. So you may disable 10 tests at a time and see what happens.
Regards
Alex