| Author |
Message
|
| HubertKleinmanns |
 Posted: Fri Sep 20, 2019 6:08 am Post subject: |
 |
|
Shaman
Joined: 24 Feb 2004
Posts: 732
Location: Germany
|
| bruce2359 wrote: |
| What a wild ride down that rabbit hole. |
| bruce2359 wrote: |
| MQI stats would be helpful here. Also a concurrent MQI trace would be helpful. |
Do you mean the values for BackCount, ExpiredMsgCount and PurgeCount? These may give a hint, but they aren't queue specific.
But now I have two new questions:
1. Are expired messages successful destructive get requests?
2. Are purged messages successful destructive get requests?
So do these increase the GetFailCount or the GetCount or none of them?
_________________
Regards
Hubert |
|
| Back to top |
|
 |
| bruce2359 |
| Posted: Fri Sep 20, 2019 1:32 pm Post subject: |
|
|
Poobah
Joined: 05 Jan 2008
Posts: 9472
Location: US: west coast, almost. Otherwise, enroute.
|
| HubertKleinmanns wrote: |
| bruce2359 wrote: |
| What a wild ride down that rabbit hole. |
| bruce2359 wrote: |
| MQI stats would be helpful here. Also a concurrent MQI trace would be helpful. |
Do you mean the values for BackCount, ExpiredMsgCount and PurgeCount? These may give a hint, but they aren't queue specific. |
Metrics offered by the OP are insufficient for adequate diagnostics. Similarly, only a car engine RPM, or oil temp, or oil pressure, or speed, or water temp, or ... any single metric at a given instance does not provide context.
_________________
I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
| Back to top |
|
|
| Andyh |
| Posted: Sat Sep 21, 2019 8:54 am Post subject: |
|
|
Master
Joined: 29 Jul 2010
Posts: 239
|
Purged or expired messages are neither successful nor failed MQGET's and wouldn't update either MQGET counter.
Imagine you were to purge a queue, then put a message with a short expiry time, and to issue an MQGET after the message had expired.
You would see one failed MQGET, and that MQGET would have completed with a 2033.
Note also that the purged counter is a count of the number of purges, not a count of the number of messages purged. |
|
| Back to top |
|
|
| bruce2359 |
| Posted: Sat Sep 21, 2019 9:46 am Post subject: |
|
|
Poobah
Joined: 05 Jan 2008
Posts: 9472
Location: US: west coast, almost. Otherwise, enroute.
|
So, to be a bit more precise, you're referring to the MQCO_DELETE_PURGE close queue option, yes?
_________________
I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
| Back to top |
|
|
| Andyh |
| Posted: Sat Sep 21, 2019 1:57 pm Post subject: |
|
|
Master
Joined: 29 Jul 2010
Posts: 239
|
No, MQCO_DELETE_PURGE relates to closing a dynamic queue, and will only update the appropriate MQCLOSE statistics.
By "purge" I'm referring to the CLEAR QL command. |
|
| Back to top |
|
|
| bruce2359 |
| Posted: Sun Sep 22, 2019 2:38 am Post subject: |
|
|
Poobah
Joined: 05 Jan 2008
Posts: 9472
Location: US: west coast, almost. Otherwise, enroute.
|
| Andyh wrote: |
No, MQCO_DELETE_PURGE relates to closing a dynamic queue, and will only update the appropriate MQCLOSE statistics.
By "purge" I'm referring to the CLEAR QL command. |
... which would imply that GetCount would include those messages "purged" by using the MQExplorer clear qlocal option to do so via MQGETs, yes?
Is it possible that other apps had the queue open, and were MQPUTting and MQGETting messages? Queue accounting data would be helpful.
_________________
I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
| Back to top |
|
|
| Andyh |
| Posted: Sun Sep 22, 2019 10:06 am Post subject: |
|
|
Master
Joined: 29 Jul 2010
Posts: 239
|
I'm not sure I understand the question, maybe I'm showing my lack of knowledge where MQExplorer is concerned ?
Removing messages from a queue using MQGET is not equivalent to issuing a runmqsc CLEAR QL command. A CLEAR QL would fail if any locked messages existed, for example an in-tran MQPUT or an in-tran MQGET, while a simple destructive GET would be unaware of any such messages.
A CLEAR QL will also clean up the resources owned by the queue better than a simple sequence of MQGET's, for example a CLEAR QL will truncate the underlying file system object associated with the queue, where a simple sequence of MQGET's will not. |
|
| Back to top |
|
|
| HubertKleinmanns |
| Posted: Mon Sep 23, 2019 12:04 am Post subject: |
|
|
Shaman
Joined: 24 Feb 2004
Posts: 732
Location: Germany
|
Andyh,
what's about "DELETE QL(...) PURGE"? Does this increase the PurgeCount and not affect the GetCount and GetFailCount - like "CLEAR QL(...)"?
_________________
Regards
Hubert |
|
| Back to top |
|
|
| RogerLacroix |
| Posted: Mon Sep 23, 2019 11:04 am Post subject: |
|
|
Jedi Knight
Joined: 15 May 2001
Posts: 3264
Location: London, ON Canada
|
| HubertKleinmanns wrote: |
| what's about "DELETE QL(...) PURGE"? Does this increase the PurgeCount and not affect the GetCount and GetFailCount - like "CLEAR QL(...)"? |
If you delete the queue then how do you gets statistics on it?
Regards,
Roger Lacroix
Capitalware Inc.
_________________
Capitalware: Transforming tomorrow into today.
Connected to MQ!
Twitter |
|
| Back to top |
|
|
| Andyh |
| Posted: Mon Sep 23, 2019 12:25 pm Post subject: |
|
|
Master
Joined: 29 Jul 2010
Posts: 239
|
In the case of a delete ql(...) purge then the "purge" is really just permission to delete the queue even if it's hosting messages. I don't believe this would make any change to the purge (as in CLEAR QL) counters.
In the case of stats written to the system topics then no further stats will be produced after the queue is deleted.
In the case of more traditional stats (controlled by the STATQ attributes) the statistics for a deleted queue should be included in the statistics messages written at the end of the interval in which the queue was deleted. |
|
| Back to top |
|
|
| bruce2359 |
| Posted: Mon Sep 23, 2019 5:20 pm Post subject: |
|
|
Poobah
Joined: 05 Jan 2008
Posts: 9472
Location: US: west coast, almost. Otherwise, enroute.
|
I'm no longer lonely down this rabbit hole...
_________________
I like deadlines. I like to wave as they pass by.
ב''ה
Lex Orandi, Lex Credendi, Lex Vivendi. As we Worship, So we Believe, So we Live. |
|
| Back to top |
|
|
| HubertKleinmanns |
| Posted: Mon Sep 23, 2019 9:53 pm Post subject: |
|
|
Shaman
Joined: 24 Feb 2004
Posts: 732
Location: Germany
|
| RogerLacroix wrote: |
| HubertKleinmanns wrote: |
| what's about "DELETE QL(...) PURGE"? Does this increase the PurgeCount and not affect the GetCount and GetFailCount - like "CLEAR QL(...)"? |
If you delete the queue then how do you gets statistics on it?
Regards,
Roger Lacroix
Capitalware Inc. |
I don't know, how a "DELETE ... PURGE" internally works. I could hazard a guess, that a very last statistics record could be written, just before the queue disappears  .
_________________
Regards
Hubert |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Tue Sep 24, 2019 12:38 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
Hubert: bevor = German; before=English; I took the liberty to fix your post... 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| HubertKleinmanns |
| Posted: Tue Sep 24, 2019 1:25 am Post subject: |
|
|
Shaman
Joined: 24 Feb 2004
Posts: 732
Location: Germany
|
| fjb_saper wrote: |
| Hubert: bevor = German; before=English; I took the liberty to fix your post... |
Ooh sorry  and thanks for correcting the typo 
_________________
Regards
Hubert |
|
| Back to top |
|
|
|
|
