G
Guest
I have a web app running on Windows 2000 with the home directory pointed to a
network share on a file server. We have a long running process which copies
jpg images from user's folders on a network share, processes them and copies
them to another location. Depending on the number of images this can take a
long time. If I update the dll (assembly) in the /bin folder the process dies
and the user is given a "Server Application Unavailable" page. When I look in
the event log all that is stated is that the aspnet_wp.exe stopped
unexpectedly.
I thought when a process was using an assmbly and the assembly was updated
the process continued to use the assembly (as a shadow copy) until it was
complete. Obviously, I'm mistaken.
I realize I should update during the obscure hours of the night, however
this is not the answer I am looking for because of other constraints. Is
there a fix? What are my options for a workaround? I realize I could do the
processing offline, but the clients want the files processed in realtime. Is
there a way I could trap the failure and recover from it when the worker
process restarts?
Thanks,
Mark
network share on a file server. We have a long running process which copies
jpg images from user's folders on a network share, processes them and copies
them to another location. Depending on the number of images this can take a
long time. If I update the dll (assembly) in the /bin folder the process dies
and the user is given a "Server Application Unavailable" page. When I look in
the event log all that is stated is that the aspnet_wp.exe stopped
unexpectedly.
I thought when a process was using an assmbly and the assembly was updated
the process continued to use the assembly (as a shadow copy) until it was
complete. Obviously, I'm mistaken.
I realize I should update during the obscure hours of the night, however
this is not the answer I am looking for because of other constraints. Is
there a fix? What are my options for a workaround? I realize I could do the
processing offline, but the clients want the files processed in realtime. Is
there a way I could trap the failure and recover from it when the worker
process restarts?
Thanks,
Mark