[Swift-devel] cleanup fails on abe

[wilde at mcs.anl.gov]

 If netcat blocks we should have it time out in a few seconds at most.

It would be good to see where its blocking. Im not sure where that would happen to send a UDP packet - its one reason we're using UDP.

The next time anyone sees a hung nc, can you find out what its waiting on, with some ps options like this:

  ps -o pid,wchan=WIDE-WCHAN-COLUMN -o comm

(Others may suggest a better way to do this?)

I see this note about netcat blocking:

From:  http://nc110.sourceforge.net/ 

"Where many other network apps use the FIONBIO ioctl to set non-blocking I/O
on network sockets, netcat uses straightforward blocking I/O everywhere.
This makes everything very lock-step, relying on the network and filesystem
layers to feed in data when needed.  Data read in is completely written out
before any more is fetched.  This may not be quite the right thing to do under
some OSes that don't do timed select() right, but this remains to be seen."

I dont see where it would block trying to send a UDP packet, so either there are cases where it can (eg buffers full? firewalls making a send() hang?), or maybe its also doing some kind of read/recv?

- Mike


----- "David Kelly" <dk0966 at cs.ship.edu> wrote:

> On Fri, Aug 13, 2010 at 2:37 PM, Justin M Wozniak <
> wozniak at mcs.anl.gov > wrote:
> 
> 
> 
> (Note that this now also may leave nc processes running, use killall
> if those pile up.)
> 
> 
> I couldn't get this to happen on my machine, but I suspect ctrl-c is
> somehow causing it to sit and wait for input. I just checked in a new
> version of the swift script which adds a timeout value. The process
> should die after 60 seconds. Please let me know if you're still seeing
> any hanging processes and I'll take a closer look at it.
> 
> David
> 
> 
> _______________________________________________
> Swift-devel mailing list
> Swift-devel at ci.uchicago.edu
> http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel

-- 
Michael Wilde
Computation Institute, University of Chicago
Mathematics and Computer Science Division
Argonne National Laboratory