Lengthy SNA Distribution Recovery step in IPL

[Pete Helgren <pete@xxxxxxxxxx>]

 A few weeks back I dealt with an open relay issue on a 520 that I 
eventually got sorted out.  At least I *thought* I got it sorted out.  
The biggest fallout was that there were a jillion spool files generated 
in the process and it took a dltsplf and a rclsplstg to get back some 
critical space.  But I still have my suspicions that there are still 
files lurking.  What raised my suspicions was that last night when I 
IPL'd the SNA Distribution Recovery step took over an hour.  When I had 
IPL'd the system after the rclsplstg a few weeks ago the SNA 
Distribution Recovery step also took equally as long.  Seems like SNA is 
"recovering" but recovering what?

So the questions are: Why is the IPL taking so long at that step 
(instead of minutes, like it used to)? Where should I look for 
"leftovers" after the open relay?  Could there be undelivered messages 
queued up somewhere?  WRKDSTQ shows nothing in any of the queues.

Thanks