RE: V5R2 -- SYSTEM21

We are a System 21, heavily modified 3.5.0b user (our System Manager/Machine 
Manager versions are current, however) and this past weekend successfully 
upgraded to V5R2M0 on a 730-1510 quad processor running at about 80% disk 
utilization, a 3570 tape drive using 7GB tapes.  We had previously tested this 
version on our development system after having determined that we didn't 
believe we had any SQL issues as have previously been discussed in this forum.

For those interested:

We performed an option 21 backup starting at 1:25am on Saturday and it took 
until 11:25am (10 hours and 14 tapes) to complete.  I started the OS upgrade at 
12:00 (noon) and received my first sign on at 2:30pm.  I then did the CUME CDs 
and they took 1.75 hours to process.  The HIPERs, DB and other product PTFs 
took me until 6:15pm (so, another 1.5 hours).  I performed an option 22 backup 
(SAVSYS + ) which took 30 minutes. I had written a CL to submit the STROBJCVN 
for all libraries to four job queues and they ran from 7:00pm to 9:00pm.  I did 
another IPL just to see if the system reported any problems resulting from the 
object conversions and had no problems, reset our startup program, re-IPLd the 
system at 9:30pm, called in my post-install "testers" and we were out of here 
by 11:00pm.

We did encounter three problems not seen in testing:

1)      Writers for printers configured with AFP(*YES) wouldn't start.  It 
turned out that there is a PTF to correct this issue that is new since the last 
CUME/HIPER.  It took about two hours to download this and the related PTFs (the 
first PTF takes almost 1.5 hours) through our ECS modem. Applying this 
corrected the problem.  We didn't encounter this in our testing because all of 
our writers on our development system are remote writers mapped to printers on 
the production system and we failed to test one of our AFP(*YES)-configured 
writers (though I'm not sure the problems exists with remote writers).

2)      Some (only a few) Client Access users at sign on are receiving an 
information panel titled "CWBUNPLA" on their PC with the message "Error 
occurred while retrieving usage information for [userprofile] on [systemname]." 
 The information panel has an "OK" button on it and clicking it allows the user 
to proceed to sign on.  

I reviewed the information on the www.ibm.com/eserver/iseries/support website 
and found on our system everything was started as the APAR suggested.  So, at 
this point, our remedy for these users is to upgrade to the V5R2 CA version. 
There is also a thread on this on www.midrange.com/midrange-l.  Again, it did 
not lead me to any alternate solution.

3)      We perform some EDI with suppliers using FTP over a secured VPN.  The 
FTP session for at least one supplier failed yesterday when attempting to sign 
on.  While I'm not entirely sure how I fixed the problem (because I changed 
multiple things at once), one thing I did was to authorize the supplier's user 
profile to a few libraries (contained in system value QUSRLIBL) where the 
profile had previously had *EXCLUDE authority.  It seems like the default 
library list of FTP has changed and where it didn't use to use QSYSLIBL, it now 
does - again, I'm not sure on this one. 

==Kevin Brunk
The Butler Company
5600 Blazer Parkway
Dublin, OH 43017-7545


**********************************************************************
CONFIDENTIALITY NOTICE:  The information transmitted in this message is 
intended only for the person or entity to which it is addressed and may contain 
confidential and/or privileged material.  Any review, retransmission, 
dissemination or other use of this information by persons or entities other 
than the intended recipient is prohibited.  If you received this in error, 
please contact the sender and destroy all copies of this document.  Thank you.  
The Butler Company
**********************************************************************