BizTalk 2004: File Receive Location Problem

Greetings all.

I am investigating/trying to resolve an issue with BizTalk and a
pile of file receive functions monitoring different folders on a remote
linux box via file shares..

I have noticed that the files on our file shares aren't picked up
some times and that the receive locations end up shutting down saying it
exhausted network retry events.

While scouring google groups, I came across a post by Rasmus
Kristensen (subj: Files are taking long time to be picked up by a file
receive location on a network fileshare) which talks about the redirector
object / current commands counter in PerfMon and that if the value is to
high, it will cause problems. He also mentioned setting the
"FileReceivePollingInterval" for the biztalk service to 1 to prevent
polling of the folder (which reduced the number of current commands.)

I attempted this as well but noticed that files were just sitting in
our directories, not getting picked up at all. The BizTalk receive
locations were still running however but they didn't seem to do anything.
If I stopped and started BizTalk, shortly thereafter the files would
disappear and be processed.

So, being the vaguely curious fellow that I am, I started looking
into this redirector::Current Commands counter to find out what exactly
it is. And what values I should be expecting for this counter so I have
a rough idea what to look for. I came across the following:

"Shows the number of requests to the Redirector that are currently queued
for service. If this number is much larger than the number of network
adapter cards installed in the computer, the network(s) and/or the server
(s) being accessed are seriously bottlenecked."
(source:
http://www.microsoft.com/resources/documentation/Windows/2000/server/resk
it/en-us/Default.asp?
url=/resources/documentation/Windows/2000/server/reskit/en-
us/counters/counters2_irof.asp)

And this:

"Current Commands: This counts the number of requests to the Redirector
that are currently queued for service. Ideally, this value should not be
more than one command per installed NIC. If a value higher than one is
present, the redirector may be the bottleneck. This can happen if there
is an asymmetry between the speed of server and the redirector which is
causing inefficient communication between the local and remote systems.
This can also indicate that the network capacity is being pushed to its
limits and that it may be necessary to break the network into subnetworks
to reduce traffic."
(source: http://www.windowsitlibrary.com/Content/435/09/3.html)

So, the questions then are:
1) Why is this counter so high? On our current production box, the value
is lurking at about 34 (it varies slightly from time to time) and I am
pretty sure that we do not have 34 network cards on this machine. Our
Biztalk Server has roughly 27 receive locations pointing to various
directories on this Linux box where clients FTP files to. Plus a few
other receive functions pointing to other boxes (windows 2003) and local
directoriess
a) Does this counter go up by one for every file receive location
defined?
b) Should I *really* be concerned about this counter, or am I just
focusing on the wrong issue because the counter value is abnormally high
compared to what seems to be the going theory on how high this should be?
c) What happens if we put a pile more file recieve locations
pointing to this Linux server (say 30 more). Rasmus, mentioned in the
thread I mentioned at the start of this post that there seemed to be a
ceiling of about 50 current commands. Would this start causing BizTalk
to shut down receive locations more frequently because network attempts
have been exhausted?

2) What can I do to lower this counter? Is there anything which can be
done or will it 'just be that high given the number of file receive
locations being used'? Should I change the configuration and just use
the FTP file receive to ftp files from our FTP server to BizTalk instead?
(I would rather not)

3) I have also been noticing a lot of events from the MRxSmb source
stating that "The redirector failed to determine the connection type"
(event ID: 3019. See: http://www.eventid.net/display.asp?eventid=3019
&eventno=149&source=MRxSmb&phase=1). Given that this seems to be related
to the Redirector (mentioned above) as well, does this have something to
do with this problem? I keep reading in other posts that this warning
can safely be ignored. This just doesn't sit right with me given that
the current command counter seems high according to what I am reading.

Anyone have any thoughts/ideas on the subject?

Thanks in advance..

Craig..

.

Relevant Pages

  • Re: Modelling Complex Sentences
    ... re-inventing the network data model). ... If such as classifying node does not exist, ... Since people complained that dbd's scripts were too long, ... you typically use the CREATE/INSERT commands to ...
    (comp.databases.theory)
  • Re: xcopy access denied
    ... reinstalling all network components might help. ... Type these commands and report the results. ... "Share" permissions (as opposed to the NTFS permissions) ... Volume in drive \\Server01\users is Raid 5 ...
    (microsoft.public.windowsxp.general)
  • Re: VAX 4000/500 general questions
    ... >The computer came from a company which went bust and has no manuals. ... >computer part of a non-existent DECnet network. ... >Which file is the most likely place to find the startup commands, ...
    (comp.sys.dec)
  • Re: CANT LOGIN!! (urgent)
    ... What is the response you get when you run these commands ... get to c$ drive over the network that didn't work" you need ... Your system drive letter has changed. ...
    (microsoft.public.windows.server.general)
  • Re: MRxSmb event log just started on all home computers.
    ... The redirector could not determine the physical characteristics of the ... connection, for example, the media type or bandwidth. ... So I am wondering if my home network using ADSL is coming into play ...
    (microsoft.public.windowsxp.configuration_manage)