Folder/File Availability test giving wrong results

[rasc]

Hi,

we mapped a drive U: as UNC path \\server\d$ .
for testing for Folder/File Availability of U:\DATA\UserTSProfiles.
But we (too) often get errors (Alive ratio : 98.14 %, Dead ratio: 1.86 %).

I sniffed the SMB traffic of HM (with only that single test enabled)
a) with WireShark on HM station
b) with wireshark on a separate computer mirroring
b1) the switch port of server
b2) the switch port of the HM machine

Whenever the error was thrown there was not a single SMB packet (request) in the stack (a) or on the wire (B). I.e. the server received no request.
We do not face any (other) connection issues on that machine. Neither (other) HM tests nor accessing UNC paths with Explorer


Could that be a problem with HM?

Thank you, rasc

[KS-Soft]

No, I don't think there is any mistake in HostMonitor code. HostMonitor uses pretty simple code for this test (related on Windows API). Probably bug in network client or some 3rd party software (antivirus monitor, personal firewall).
However if you tell us what exactly version of HostMonitor do you use, we will check the code.

BTW: "Bad" status for this test means HostMonitor received necessary information (response from Windows API regarding target file or folder).
Unknown status means HostMonitor did not receive invormation, Windows API returned some error code.

Regards
Alex

[rasc]

Dear Alex,

that's a pitty! It would have been much easier (for us) if there was a bug in your software
We use 8.14, are unfortunately not eligible for newer versions.

Thanks for your feedback!

[KS-Soft]

We do not see any mistake in our code.
I assume you do not use Remote Monitoring Agent (RMA) for this test?
I would recommend to use UNC path instead of mapped drive, especially if HostMonitor is started as service.

Also, what exactly is "bad" test status? Bad? Warning? Do you use "Optional status processing" options for this test?

Regards
Alex

[rasc]

Hello Alex,

- when using HM to check we get these errors. Most of the time its just 1 consecutive 'Bad' event.
- 'Bad' means that your software says 'status bad'. Whatever the deeper causes are, I just see your software saying 'status: bad' (not unknown) in the 'mail to admin'.*
OSP settings are 'treat warning as bad' only, no other checked.
- using the agent on the server brings (almost) no error
- using the suggested UNC test brings slightly less errors


*
Message from HostMonitor (host changed status)

Test : UserTSProfiles available
Method: File exists
Status : Bad
Date : 24.11.2009 14:10:33
Reply :

Recurrences : 2
Last status: Bad
Total tests: 19
Alive ratio : 89.47 %
Dead ratio: 10.53 %
Comment: U:\DATA\UserTSProfiles
Folder: Server

[KS-Soft]

Well, Bad status for this test means Windows API client "said" everything is fine and provided intormation about target file/folder.
May be you are using "Alert when file doesn't exist or older/newer than N min" test option?

However this cannot explain why system does not send request to remote system May be it does?
Perhaps this problem (as many other mystical problems) relates to some buggy antivirus software? Do you use any antivirus monitor, personal firewall, non standard winsock components? Could you disable such software?

Regards
Alex

[rasc]

Dear Alex,

I am sorry for the late reply.
We just skipped that test as the effort to get it reliably working seems much to big to us.
I am sorry about that to as it's one chance less to further improve your products.

Best regards

[KS-Soft]

No problem.
Anyway, we do not thing there is any mistake in our code. This test uses very simply code and it works for many years without problems. May be there is some bug in network client...

Regards
Alex