Richard Chandler
2004-02-17 12:32:54 UTC
Hi All,
I have a very strange problem and seems to contradict the laws of COM/.Net
and asychronous comms.
I have a scanner that works fine asynchronously on 99.5% of computers. It's
written in C#.
However on a few machines (1 in 500) we experience a problem whereby the
asynchronous Get(...) method of System.Management.ManagementObjectSearcher
locks up trying to initiate the asychronous comms. This leaves my scanner
with some threadpool threads locked up inside this Get(...) call.
I had one remote machine that waited over 20 minutes before giving up
telling me that the problem was something to do with the RPC services (I'm
currently contacting the customer to get the exact HRESULT), but not that
they weren't registered or the machine wasn't available ... something was
keeping it talking but not asynchronously. The others I gave up waiting for
a response from.
But I thought that COM/RPC would give up much sooner than this anyway if the
device wasn't reachable or something was wrong, so is this the WMI subsystem
not fully working
Wbemtest has exactly the same problem on the same remote machine, either
asynchronous or not. I haven't tried wbemtest locally on the remote machine.
The remote machine is an NT 4 box... I'm trying to get more info on the
problem machines and the exact error number ... sorry to be a bit vague at
the moment.
Has anyone come across a problem like this before ... any articles ...
threads ... anywhere ?
Thanks,
Richard
I have a very strange problem and seems to contradict the laws of COM/.Net
and asychronous comms.
I have a scanner that works fine asynchronously on 99.5% of computers. It's
written in C#.
However on a few machines (1 in 500) we experience a problem whereby the
asynchronous Get(...) method of System.Management.ManagementObjectSearcher
locks up trying to initiate the asychronous comms. This leaves my scanner
with some threadpool threads locked up inside this Get(...) call.
I had one remote machine that waited over 20 minutes before giving up
telling me that the problem was something to do with the RPC services (I'm
currently contacting the customer to get the exact HRESULT), but not that
they weren't registered or the machine wasn't available ... something was
keeping it talking but not asynchronously. The others I gave up waiting for
a response from.
But I thought that COM/RPC would give up much sooner than this anyway if the
device wasn't reachable or something was wrong, so is this the WMI subsystem
not fully working
Wbemtest has exactly the same problem on the same remote machine, either
asynchronous or not. I haven't tried wbemtest locally on the remote machine.
The remote machine is an NT 4 box... I'm trying to get more info on the
problem machines and the exact error number ... sorry to be a bit vague at
the moment.
Has anyone come across a problem like this before ... any articles ...
threads ... anywhere ?
Thanks,
Richard