Monitoring AQ instances

What is a good way to monitor AQ instances? I was looking at using GV$AQ and/or
V$AQ views;
Worry with GV$AQ is that it does cross instance traffic and may contribute to
interconnect traffic;
Ran a trace on V$AQ and it does not appear to get it's stuff from
the SGA, but does runs queries on the queue tables themselves. Based on the
sql traces, it looks like these might get expensive. (there are full table scans in
the plans)
Any opinions? I am in a 9.2.0.7 version, 4 node environment with one queue with affinity to each rac node.
Thanks
Vijay

Here are more stuff now in 10g
New DBMS_AQ packages
New DBMS_AQADM packages
V$BUFFERED_QUEUES
V$BUFFERED_SUBSCRIBERS
V$BUFFERED_PUBLISHERS
--Khaleel

Similar Messages

Issue monitoring SQL instance

We are running SCOM 2012 R2 and are monitoring several SQL 2012 servers.
There is one I am having difficulty monitoring. Within the event logs I am constantly getting:
Management Group: SCOM Script GetSQL2012DBFileGroupFreeSpace.vbs : Cannot login to database [servername][mssqlserver:content_portal]
For all the SQL databases that are on the server.
I am using a dedicated Action Account that is a sysadmin on the SQL instance, a local admin on the server and I have configured the Default action profile to use the dedicated action account for that server.
I have also ran the following command:
USE [master]
GO
/****** Add a login in SQL Server for the service SID of System Center Advisor HealthService ******/
CREATE LOGIN [NT SERVICE\HealthService] FROM WINDOWS WITH DEFAULT_DATABASE=[master], DEFAULT_LANGUAGE=[us_english]
GO
/****** Add the HealthService Service SID login to the sysadmin server role ******/
ALTER SERVER ROLE [sysadmin] ADD MEMBER [NT SERVICE\HealthService]
GO
Restarted the Microsoft Monitoring Agent several times but still the cannot login to database error.
The above setup works fine for the other two SQL 2012 servers we are monitoring.
Am I missing something?
www.techielass.com

have a read at the below
http://www.scomgod.com/?p=258
Extract of the needed permissions is as below
For Low Privilege environments:
o   I add the Run As account or Action Account to:
o   Local Performance Monitor Users Group
o   Local EvenLogReaders Group
o   Add Log On Locally permission
o   Configure the SQL Instance for monitoring
Details and Permission required are stated in Low Privilege Environment monitoring in the MP Guide
Note from the MP Guide:
“By default, all discoveries, monitors, and tasks defined in the SQL Server management packs default to using the accounts defined in the “Default Action Account” Run As profile. If the default action account for a given system does not have the necessary permissions
to discover or monitor the instance of SQL Server, then those systems can be bound to more specific credentials in the SQL Server Run As profiles, which do have access.”
Blog: http://theinfraguys.com
Follow me at Facebook
The Infra Guys Facebook Page
Please remember to click Mark as Answer on the answer if it helps you in anyway

Need help in Solution Manager System Monitoring-activating Instance Status

I would like to use System Monitoring with Solution Manager.
The agents CCMSPING and SAPCCM4X are installed on the monitored system and CCMS is configured
I've already created a solution and a logical component. Now i would like to activate Instance Status in the tree when i enter in Setup System Monitoring .... but i get a red alert an the following Message Instance status is no performance alert. Thresholds have been deleted ... how can i activate this Instance Status?
I am going my first steps in SAP SolMan and tried to solve this problem but i had no success. Can someone please help me? What i am doing wrong?
thanks...
Edited by: Satish Machavarapu on Dec 15, 2009 12:46 PM
Edited by: Satish Machavarapu on Dec 15, 2009 1:04 PM

Hi,
If you want to monitor the Availability status of the instance then check the below docs
[http://help.sap.com/saphelp_nw70/helpdata/EN/38/9e100dd7980fffe10000009b3810a5/frameset.htm]
[CCMSPING config|http://www.sdn.sap.com/irj/scn/elearn?rid=/library/uuid/d547f687-0901-0010-07a9-ce64cc044215]
[CCMS Agent config|http://www.sdn.sap.com/irj/scn/elearn?rid=/library/uuid/40ef0688-814a-2a10-afb1-89b40d83d0e0]
regards
Naveen
Edited by: Naveen Kumar on Dec 15, 2009 1:07 PM

Monitor an instance's changes

Hello,
all information that I need to monitor with a unit monitor is included in the properties of its target instance. The monitor must react to changes of the properties of its target instances (as they change during discoveries). Such as, if one of its properties
changes from 1 to 2, monitor should go critical, when the instance's property changes from 2 to 1, it should switch back to green.
Is it possible? To have a monitor receive both the current and "yet uncommitted future" state of an instance which it targets?
ondrej.

hello, thanks for your interest, but it seems I do not understand. I am kind of new into MP authoring. So would you please be able to be more specific?
My class instance and its parameters are discovered with a very resource expensive operation. It takes something like several-minutes-on-one-CPU-core-at-80% to populate the properties of the instance. I do that discovery with a script surely.
If I want to monitor the changes, I can definitely use some similar script for the "monitor" to plug data from the physical system again. The problem is, I just want to save some resources and do not want to run the script again.
The resulting value is already saved in the instance as one of its properties.
So isn't there a way, a monitor or something, that would be able to react/target directly the changes in a discovered instance?
I am just inexperienced in this. What I can do now, is to extract some monitor input values from its $Target, but these are only the current properties of the instance. I need to see even the previous values after its change.
I need to react NOT ONLY to a specific constant property value. I need to see the previous and current property states to determine how to react.
Is there a way how one can create a monitor/event/anything that would receive as parameters previous/future properties of some instance?
ondrej.

Monitoring failed instances

Dears,
Do we have a possibility to monitor the scheduled webi reports?
- Could we inform a dedicated user with an email, if an instance failed
- Could we see in the CMC a list of all failed instance e.g. in the CMC
As our customer has scheduled dozens of reports, its not applicable every report manual. The Instance manager only shows the scheduled jobs but not the Status. Thanks for your feedback.
SAP BI Platform 4.1 SP2

Hi Felix,
The instance manager in the CMC is better than you think.
You can search for Failed instances - and if there is one you can click on failed and get the reason for the failure
Another option that you have is to automate the email writing and use an scheduling event. Please see the next three screenshots to setup an event and also the last screenshot where to define it in the publication.
and finally the setting in the properties
Best regards
Manfred

Monitor SAP Instance using Webdynpro

Experts,
I have a requirement to show the server SAP Instance status in WebDynpro Application.
for an example:
My server name : myserver
SAP Instance : DE3
I need to create a webdynpro iView which shows the server status like below:
a) If the instance is up and running,I need to show that the server status as green.
b) If the instance is not running,I need to show that the server status as red.
Kindly let me know how to proceed.
Thx in Advance.
Regards,
Subbareddy

hi sdnuser1...
you can create such web dynpro application....
you have tutorial for service call..
create aservice call in such a way that it displays the po in a table....
have the table in edit mode....
the values would be present in an internal table...
so once the values are changed have an update button.
call the respective bapi to update it in the database.
---regards,
alex b justin

Monitor Oracle instance in Enterprise Manger

Hello,
My Oracle database server is runing on Windows system. The version is Oracle 10.2.0.4 standard edition. I tried to set up the email alert when the database, listener goes down in Enterprise Manger.
The problem is I cannot receive the alerts which I am supposed to receive when the database and listener are down However, I can get these alerts immediately when I bring up the database. That does not make sense to me. Please help and let me know how I can get the alert when these critcal services are down from Enterprise Manger.
Thanks!

Need to check does your agent uploading XML files? some cases it fails to upload then you need to upload manually
emctl upload agentor reconfigure if any problem occurs still

Monitor dynamic and arbitrary performance instances with single monitor?

I need to monitor a performance object in such a way that each *instance* within that object is monitored and alerted separately with regards to thresholds. So, for example, if I have a performance object called App1, and that object has five instances (Instance1...Instance5),
I want to be able to monitor each instance separately with regards to thresholds.
The issue is that the instances are dynamic and arbitrary. They are created and deleted on the fly by a different application, and I will have no advance warning of them being created or deleted.
I'm running SCOM 2012 R2. It seems that SCOM performance monitors are missing a checkbox that says something like 'Treat each instance separately' or I'm missing how to achieve that functionality.
For reference, I'm moderately handy with VSAE if this can be hacked together that way. Note that the rate of change of these instances is relatively low, such that a daily discovery would be sufficient if that would somehow assist in enabling this functionality.
However, what I don't want to do is to have to create a separate monitor for each instance - is this possible?

It can be done only with alert rules.
Vladimir Zelenov | http://systemcenter4all.wordpress.com

Instance size logging / monitoring

I am trying to debug an instance size issue and would like to log/monitor the instance size during process execution.
I wanted to simply logMessage it and monitor it in Studio but cannot find an appropriate API in the Fuego Catalog, something like...
logMessage("current instance size = " + Fuego.Papi.Instance.size); Or something similar.
But no such size attribute exists.
Any recommendations as to the best way to achieve this logging/monitoring?
Thanks.

Hi Michael,
There's an engine setting (both on Studio and Enterprise) for the maximum allowable size of instance variables.
You have a few choices:
1. you can bump the max instance size setting on Studio and Enterprise (not a fan of this)
2. you can change the offending instance variable(s) to be "Separated" or
3. you can look at your variable and see if it contains information rarely accessed and possibly should be retrieved from a database.
good to hear from you,
Dan

How do I write/use performance monitor to run an action when something happens?

I notice that I get a popup if desktop resources are over-used asking me if I want to turn off the glass desktop.
That's fine.
I've been looking in Performance Monitor for how this is done, as I have a program that periodically runs amok and starts indexing my Documents directory. I usually start to notice my machine getting sluggish when the process gets to about 1.5-2+GB
in Virtual and working memory. At that point it seems to be increasing it's network usage rate geometically, and interrups will start getting delayed or lost -- keyboard will lock up for periods, lots of things freezing temporarily (music will get 10-20
second dropouts) It can even cause a complete system lockup
So what I would like to do is monitor any instance of this process by name ('explorer'), and restart it when it gets to over 1.5GB of usage.
I setup a monitor of virtual size and working memory in the performance monitor, but I don't see how to set up an alarm that starts an action to restart it. (restarting it will cause the problem to go away for anywhere from a few hours to several days).
I've no clue as to what triggers it.
I found a bunch of XML scripts in
Windows/pla/Rules with names like:
Rules.System.CPU.xml*
Rules.System.Common.xml*
Rules.System.Configuration.xml*
Rules.System.Diagnostics.xml*
Rules.System.Disk.xml*
Rules.System.Finale.xml*
Rules.System.Memory.xml*
Rules.System.NetDiagFramework.xml*
Rules.System.NetTrace.xml*
Rules.System.Network.xml*
Rules.System.Performance.xml*
Rules.System.Performance.xml.orig*
Rules.System.Summary.xml*
Rules.System.Wired.xml*
Rules.System.Wireless.xml*
These have some XML language in them that looks like it woud be for monitorying and doing actions:
<Rule name="$(RuleDiskSummary)" enabled="true">
   <Step select="/Report/Section/Table[@name='hotFile']">
    <Exists>
     <Otherwise>
      <Delete select="Summary"/>
     </Otherwise>
    </Exists>
   </Step>
   <Step select="/Report/Section/Table[@name='physicalDiskCounters']" fatal="true">
    <Exists>
     <Otherwise>
      <Insert>
       <Node>
        <Summary key="100" find="field" field="counter" value="Avg. Disk Queue Length" topic="diskInfo">
         <Data name="max" label="diskQueue"/>
        </Summary>
       </Node>
      </Insert>
     </Otherwise>
    </Exists>
   </Step>
</Rule>
</Group>
<StringTable>
<String ID="GroupSummary">Add Summary Blocks</String>
<String ID="RuleMemorySummary">Memory Summary</String>
<String ID="RuleDiskSummary">Disk Summary</String>
<String ID="RuleNetworkSummary">Network Summary</String>
</StringTable>
<Rule name="$(RuleDiskSummary)" enabled="true">
   <Step select="/Report/Section/Table[@name='hotFile']">
    <Exists>
     <Otherwise>
      <Delete select="Summary"/>
     </Otherwise>
    </Exists>
   </Step>
   <Step select="/Report/Section/Table[@name='physicalDiskCounters']" fatal="true">
    <Exists>
     <Otherwise>
      <Insert>
       <Node>
        <Summary key="100" find="field" field="counter" value="Avg. Disk Queue Length" topic="diskInfo">
         <Data name="max" label="diskQueue"/>
        </Summary>
       </Node>
      </Insert>
     </Otherwise>
    </Exists>
   </Step>
</Rule>
</Group>
<StringTable>
<String ID="GroupSummary">Add Summary Blocks</String>
<String ID="RuleMemorySummary">Memory Summary</String>
<String ID="RuleDiskSummary">Disk Summary</String>
<String ID="RuleNetworkSummary">Network Summary</String>
</StringTable>
Question is, how can I setup it for calling a trigger script when
parameters for Explorer get out of spec.
On confusing issue that stopped me from proceeding, was the time schedule to run the report:
I only allows for once/day.
I need / want continuous monitorying -- say at 1 minute or 3 minute intervals (5 might work, not sure -- doesn't take long for Explorer to get to 1-2G with a network connection of 10Gb.
So how can I monitor and control it?
A side issue -- I notice a resource-control agent for Win 2008 -- but this was disabled for Win7.
Anyone know why resource usage on Win7 was considered less important than on Win2k8 server?
My version of Win7 is Win7-x64-sp1 "Ultimate" [sic] (it doesn't include professional features like resource control)...
Thanks...This has bothered me for over a year and has persisted through new profiles and reinstalls of of the OS.
I have a feeling it is related to MS disabling Network disks in their indexer, so Explorer ends up doing it as a surrogage for things like the search engine (with the results not stored, because they are not local -- the worst of both worlds?)... but that's
just a feeling...
Thanks!
Astara...

You can create a Data Collector Set (create manually, not template), then choose performance counter Alert, configure alert actions, from the script, seems you want to monitor the “physicalDisk” performance, so you can set the “physicalDiskCounters”
into alert action, please follow the instruction in the link below.
Monitor disk space
http://www.ms-opsmgr.eu/?cat=43
If you want to view the detailed process related about explorer when your machine getting sluggish, I'd like to introduce a tool named process explorer to you.
http://technet.microsoft.com/en-in/sysinternals/bb896653.aspx
It can clearly show you the processes running in your system, along with the hierarchical relationship and the cpu usage.
NOTE
This
response contains a reference to a third party World Wide Web site. Microsoft is providing this information as a convenience to you.
Microsoft
does not control these sites and has not tested any software or information found on these sites.
Regards
Yolanda
TechNet Community Support

Calibrating my Monitor question

Hello everyone
I am trying to calibrate my JVC monitor
i am following the instructions to a link mentioned in one of the threads
i generated a color bar in finalcut and brought it to the timeline
i see it on my monitor, i turned the chroma all the way down and the contrast to the middle
I Canot see these three lines on my monitor-
Three Picture Lineup Generating Equipment bars on the bottom right called PLuGE bars.
any suggestions
Danny

Great article. I copied the section relevant to my situation below:
Accurate Video Monitoring
To ensure accurate video monitoring, your video monitor should be correctly calibrated to your system. The key thing to remember is that the “Brightness” control on your monitor should more correctly be labeled “Black Level”. The setting of the black level control on your monitor will determine which analogue voltage you the viewer will perceive as black when you view the image on the monitor. By adjusting this control correctly you can calibrate your monitor to show you a true image whether your DV deck, for instance, adds 7.5 IRE setup correctly or not to it’s analogue outputs.
Problems can occur when you monitor multiple video sources through the same monitor, for instance an analogue deck and a digital deck. In this situation, if the digital deck does not correctly add 7.5 IRE setup, you would need two different monitor calibrations, which is tricky to achieve. The solution is to add a “proc amp”, an analogue video manipulation device that can raise the analogue output from the digital deck by 7.5 IRE, making up for the lack of such circuitry inside the deck itself.
I have a JVC deck that DOES correctly add 7.5 IRE Setup if needed, but this ONLY matters if I'm dubbing to an analogue format, correct? If I'm just monitoring through a broadcast monitor to color correct in Final Cut Pro, I should:
-use the 7.5 IRE Setup switch on the JVC deck
-calibrate the monitor to 7.5 IRE being the darkest black
My question still is, what is the 7.5 IRE Setup switch on the monitor for? Should both switches be on, or does that add 7.5 IRE twice? My gut is that the monitor 7.5 IRE switch also adds 7.5 IRE and thus should be off, provided the JVC deck 7.5 IRE switch is on.
Is anyone following me?
Thanks.

How to deploy multipe instances of Oracle Adapter

Hi,
I just started working on Oracle fusion stack.
We are integrating OracleApplications e-Business suite with PSFT. Data sync requirement.
I could configure and use OracleApps Adapter. But i am facing issues with deployment
How can I deploy & monitor two instances of Oracle adpater. one for 11510 & other for 1155?
From Oracle enterprise manager I can see only one instance available for start and stop.
I modified <oracleASHome>j2ee\home\application-deployments\default\AppsAdapter\oc4j-ra.xml to include two <connection factory> entries for 11510 & 1155. But that doesnt help to show two separate instances.
Also, why do i need to stop Oracle AS server after new adapter(psft) installation?
Oracle Patch readme for psft adapter says "bounce apps server" after adapter patching & modifying server.xml. Is that really needed?? i am really confused with this
Thanks in advance

You need to add more information than that in order to get an answer...
Now I'm guessing... You're deploying Java applications to an OC4J. In this case, create a OC4J for each application you're deploying. This means that they can be individual controlled and also configured with different memory params etc.
Regards,
Martin

What would cause a lot of redo or logswitches in an instance with no users?

Hello all,
I've got a database instance that is set up as a 3 node RAC cluster. I don't think this is RAC related, but the database issues is causing RAC issues.
The instance has been set up, and later will be used to temporarily hold data while another box is being rebuilt. This instance has not had anything imported into it...there are no users connecting to it.
However, it seems to be generating a LOT of redo and archive log traffic. So much so..that overran the db_recovery_file_dest_size of 8GB.
I'd not been monitoring this instance like I said before....nothing is in it, and noone should be connecting to it...but it filled up the 8G..and then, started writing tons of logs to /u01/app/oracle/diag/rdbms/instance1/INSTANCE1_node3/trace/ saying it was full...and eventually filling up the /u01 filesystem..causing things to crash.
I fixed all this by deleting the alert logs to make room. I then did an RMAN job to move the archive logs to tape..and it came back up.
However, just an hour later...I see over 15% of the FRA is filled up again?!?!?
Any ideas on where to start to troubleshoot this? I don't know what is causing this thing to generate redo at all.....
Thanks in advance,
cayenne

I've got a database instance that is set up as a 3 node RAC clusterThat would mean 1 database with 3 instances.
Are all 3 instances writing a high volume of redo ? Or is it only one of the three instances ?
When you check for scheduled jobs in DBA_JOBS and DBA_SCHEDULER_JOBS, check all 3 instances.
Hemant K Chitale
http://hemantoracledba.blogspot.com

Colors wrong on monitor in Lightroom 3

Everything was working Ok until I updated my motherboard - Gigabyte motherboards with Intel chipsets in before and after case. The new board has an Intel HD Graphics 3000 chipset. Latest driver installed. Monitor is a Samsung SyncMaster - latest driver installed. Windows XP SP3. Display setting 32 bit at 1680 x 1050 (to match the screen resolution). Spyder3Express calibration used - latest version.
The colors are wrong on the monitor in Lightroom 3.5. The colors are OK in Windows Viewer or Canon ZoomBrowser EX. I can even export from Lightroom to a jpg and get the same colors in the jpg as I did with an export before the upgrade. The color change in Lightroom is applicable to CR2, jpg and dng files imported both before and after the motherboard upgrade.
I've updated all drivers to latest versions and uninstalled/reinstalled the Spyder calibration software, recalibrated the monitor and rebooted. The color problem still persists in Lightroom.
I suspect a bad color profile is being loaded into Lightroom but do not know how to determine the profile loaded or even where the profiles are loaded from. Any help in troubleshooting would be appreciated.

Just want to clarify a point here that many people miss, because it's not obvious. Monitor calibration and profiling are separate processes, although both are usually performed at the same time and therefore often lumped together in the word "calibration".
You see a color shift when you boot up the system, right? Many people assume that's the monitor profile, but actually it's only a small part of it. That's just a basic correction of white point and gamma curves, and this is the calibration part. Since it's loaded into the video card (or monitor) it affects everything, system-wide.
But the monitor profile is more complex. It is a full and complete description of the monitor's behaviour in three-dimensional color space, in its calibrated state. The precision is much higher. Calibration can't differentiate between a wide gamut and a standard gamut monitor for instance, but the profile will, because it pinpoints the position of the three primaries. Just to give you an idea of the difference.
So if the profile is bad Lightroom will get the wrong picture and be thrown off. The other applications live in blissful ignorance and for them it's business as usual.

Monitoring Schema Username

I am running OEM GC 10.5 with AMP 3.1. I have discovered and can monitor my various databases, PROD, DEV, TEST, etc... I am having a bit of a problem understanding the process of a discovery in order to monitor e-business through AMP 3.1.
I will begin by describing the actual problem: So I discover e-biz and always hesitate at the prompt for "Monitoring Schema Username" and "Monitoring Schema Password". I have read the AMP 3.1 users guide, but have found little mention of what the monitor schema username should be. What I have tried and used successfully during discover process is my apps user and password, my system user and password have also worked. Once e-biz is available as a target under the "Oracle Applications" tab, I can interact with tabs Home, Performance, Administration, Maintenance, and Topology. I cannot interact with Diagnostics and User Monitoring tabs. I get "Unexpected Error occured while loading the data for the screen. Refer the log for more details", and this is part of what I am getting in emoms.trc if I click User Monitoring lets say -
"2010-08-11 10:59:16,980 [EMUI_10_59_16_/console/oam/oamHome$ctxType=oamFolder$selTab=5$target=PROD-Oracle_20E-Business_20Suite$type=oracle*_ebs] WARN pref.SubtabPref getFolders.710 - Unknown folder id: VirtualServers retrieved from repository
2010-08-11 10:59:17,503 [EMUI_10_59_16_/console/oam/oamHome$ctxType=oamFolder$selTab=5$target=PROD-Oracle_20E-Business_20Suite$type=oracle*_ebs] ERROR rest.ServiceClient retryAuthentication.503 - Authentication/Authorization failure message =
2010-08-11 10:59:17,505 [EMUI_10_59_16_/console/oam/oamHome$ctxType=oamFolder$selTab=5$target=PROD-Oracle_20E-Business_20Suite$type=oracle*_ebs] ERROR handlers.DashboardHandler loadData.386 - Invocation Exception thrown
oracle.apps.fnd.oam.em.rest.exception.AuthorizationException: Unable to authenticate/authorize. The credential provided is wrong or the user doesn't have sufficient privilege.".
One thing I should mention is that I know the following to be true. I have set preferred credentials for e-biz.
Applsys Schema Username (apps)
Applsys Schema Password (my password for apps)
Apps Schema Username (apps, is this same user as above?)
Apps Schema Password (my password for apps)
System Schema Username (system)
System Schema Password (my password)
Applications Login Username (sysadmin)
Applications Login Password (sysadmin password)
Host Username (root)
Host Password (root password)
Run as (not sure what I am supposed to put here)
Profile (not sure what goes here)
So my questions are:
What should the "Monitoring Schema Username" be?
Does the correct user have to be set up in e-biz? special permission? db permissions?

Typically for monitoring purposes, you can use dbsnmp. The only exception I've seen is when you monitor standby instances, since instance is only in mounted mode, you have to use 'sys'. For applications, I suspect you would use an "admin" type of id at the app level.

Monitoring AQ instances

Similar Messages

Maybe you are looking for