I found the cause!!!

Somewhere in the exec, I have a call to NETSTAT command.  It appears
that if a message arrives at just the right time, while I am doing
the NETSTAT, I lose the message.

Is this expected?

Aria.
On Wed, 15 Feb 2006 09:44:20 -0500 Colleen Brown said:
>
>Aria,
>        The first time you invoke WAKEUP with the IUCVMSG option, you will
>load WAKEUP's IUCV interrupt handler.  It is always a good Idea to invoke
>WAKEUP once outside your loop with something like WAKEUP +00 (IUCVMSG  .
>This will wake up immediately and will load the IUCVMSG handler.  Then,
>when you get into your loop, you should be all set.
>        Also, you can set SMSG ON or set SMSG to IUCV.  WAKEUP can handle
>both VMCF and IUCV interrupts.  When you set SMSG to IUCV I believe you
>only need the IUCVMSG option.  Also, NOEXT is not needed anymore.
>
>Colleen M Brown
>IBM z/VM and Related Products Development and Service
>
>
>
>Aria Bamdad <[EMAIL PROTECTED]>
>Sent by: VM/ESA and z/VM Discussions <[email protected]>
>02/15/2006 09:12 AM
>Please respond to
>VM/ESA and z/VM Discussions
>
>
>To
>[email protected]
>cc
>
>Subject
>Re: Problem with WAKEUP
>
>
>
>
>
>
>I should have been more clear..  I am not doing the SET SMSG within the
>loop.  Here is what I have
>
>'CP SET SMSG IUCV'
>do forever
>   "WAKEUP +00:10 ( SMSG NOEXT IUCVMSG"
>   <do some work based on why we woke up>
>end
>
>There are no other SET SMSG commands in the code.  Allthe work done is
>in the same exec and do not involve RACF, etc.
>
>I do issue a new WAKEUP every time the loop is repeated because I need to
>sleep for at most 10 seconds.  I don't see any other way to do with
>without having an external wakeup times file.  Would issuing the WAKEUP
>command flush out any pending SMSG messages?
>
>Aria.
>
>
>On Wed, 15 Feb 2006 06:43:29 +0100 Kris Buelens said:
>>1. There is no need for the CP SET SMSG IUCV.  Our RXSERVER doesn't use
>it
>>2. If WAKEUP is reset in the loop: between the CP SET and WAKEUP, SMSG's
>>won't be handled
>>3. You don't say what you do every 10 minutes.  RAC(F) commands for
>>example also use VMCF, and that "kills" any pending SMSG
>>
>>Kris,
>>IBM Belgium, VM customer support
>>
>>> Hi,
>>
>>> I am having a problem with the WAKEUP command missing SMSG messages
>>> when the messages arrive just before a scheduled wakeup time.
>>
>>> I have suspected this for years but never had a chance to look at
>>> it closely.  I am now able to reproduce the problem.  Here is what
>>> I have in a loop on a DVM:
>>
>>> 'CP SET SMSG IUCV'
>>> "WAKEUP +00:10 ( SMSG NOEXT IUCVMSG"
>>
>>> The machine is to do something every 10 seconds or when it receives
>>> an SMSG.  If the SMSG arrives **JUST** before the 10 second timer is
>up,
>>> it will be missed!
>>
>>> Has anyone else seen this problem?
>>
>>> Thanks in advance.
>>> Aria
>

Reply via email to