T.R | Title | User | Personal Name | Date | Lines |
---|
3556.1 | event rate? | TOOK::PURRETTA | | Thu Aug 13 1992 17:23 | 10 |
| Gary,
I'm going to try and reproduce the situation locally and see
what's going on.
Do you have any filtering at the outbound stream end?
(PASS,BLOCK,IGNORE). Can you give me some idea of the event
rate on each of the streams?
John
|
3556.2 | trying to install filters on event streams | SEDSWS::MALLOY | | Fri Aug 14 1992 08:36 | 16 |
| John
At present the event streams are open (No blocks).
We are trying to put filter on at present by the WANrouter
are very flaky .
Hopefully with good wind we should install some filter in the near
future. I only HDLC LINK xxx Logical Station xxxx station running
and station inoperative events.
The WAnrouter seem to have a problem understanding Logical stream
in the pass global filter command.
I am on holiday for two weeks now .
Gary
|
3556.3 | Check SYSGEN parameters | TOOK::PURRETTA | | Fri Aug 14 1992 12:53 | 9 |
| Gary,
when you get back, I would suggest looking at some SYSGEN parameters
and see if they are high enough for the kind of I/O you're doing.
I would look at PQL_DBIOLM and PQL_DBYLM
John
|
3556.4 | Ithought Pqlxxx wrer used by RUN/DETACH | SEDSWS::MALLOY | | Fri Aug 14 1992 14:46 | 23 |
| John
Today, I tried to start up the MCC collector,the same problem occured.
DECmcc hang when trying to disable the mcc_sink object.
I enable MCC collector and the process was created but NO object.
REboot machine both the DNA5 and Collector object were created.
This may be a WAVE1 problem
I will check these parameter when I come back form hols
John I thought that the DNA5 process took is quotas from the process
that created it. Is ths correct ????
Run/detach , I thought used the PQLXXXX parameters
Gary
|
3556.5 | It _is_ a detatched process | TOOK::PURRETTA | | Fri Aug 14 1992 18:27 | 24 |
| Gary,
When you give the command
MCC> Enable MCC 0 DNA5_AM Sink
the DNA5 AM does a SYS$CREPRC with the PRC$M_DETACH flag set.
So the EVL sink program is created as a detached process.
That's why I suggested looking into those SYSGEN parameters.
As for the problem of enabling and not seeing a process generated
one thing I would do is a $SHOW SYSTEM and look for a process name
called MCC_DNA5_EVL. The program makes sure that it is the only one
running by making sure there isn't any other process using the same
name. If you have run the EVL interactively, then it changed "your"
interactive process name to this and the real MCC sink won't come up
until it sees it's the only one on the system with that name.
Unfortunately, this error doesn't get propogated back to the
Director. You could do an "Enable" and it would say it succeeded
but what that means is that the CREPRC system call succeeded.
Once the program is out there running on its own, it could bring
itself down for the reason I explained above.
-- John
|
3556.6 | sysgen parameter required | SEDSWS::MALLOY | | Thu Sep 03 1992 08:26 | 7 |
|
Here are the sysgen you rquired
Pql_dbiolm = 18 default
Pql_dbytlm = 8192 default
Gary
|