Event ID 2115 A Bind Data Source in Management Group

re: Event ID 2115 A Bind Data Source in Management Group

Rich — Wed, 27 Aug 2008 16:35:38 GMT

I think you missed

Microsoft.SystemCenter.CollectPerformanceData

from your workflows.

re: Event ID 2115 A Bind Data Source in Management Group

kevinhol — Wed, 27 Aug 2008 17:06:51 GMT

You are correct - I did. This was a copy/paste from a newsgroup posting.

re: Event ID 2115 A Bind Data Source in Management Group

mkielman — Tue, 23 Sep 2008 21:22:53 GMT

Kevin -

I am experiencing the problems you describe for the CollectEventData workflows. If I follow your workaround, will I be prevent specific types of event collection? Can you provide more detail in what is causing this?

Thanks!

Megan

re: Event ID 2115 A Bind Data Source in Management Group

mkielman — Tue, 23 Sep 2008 21:31:57 GMT

One other question. Can this problem cause agents to receive the event "Alert generated by Send Queue % Used Threshold"?

re: If I follow your workaround, will I be prevent specific types of event collection?

kevinhol — Tue, 23 Sep 2008 23:37:18 GMT

No - we will simply drop batches of events that get stuck and hold up the queue.

re: Can this problem cause agents to receive the event "Alert generated by Send Queue % Used Threshold"?

kevinhol — Tue, 23 Sep 2008 23:38:43 GMT

I believe so - if the management server queue is also blocked. One customer I worked with had a lot of send queue % alerts.... and these cleared up when we implemented this change.

re: Event ID 2115 A Bind Data Source in Management Group

mkielman — Wed, 24 Sep 2008 18:14:02 GMT

Thank you! It appears this fix has cleared up my issue as well including the Send Queue Alerts. Thanks!

Event ID 2115

sb — Thu, 16 Oct 2008 12:39:14 GMT

i have applied all the 3 overrides :

- Max Execution Attempt Count to 10

- Execution Attempt Timeout Interval Seconds to 6

- Max Batches To Process Before Maintenance Count to 50

but 2115 warning on my RMS server are still arriving. note that i have 2 physical rms and a virtual rms server (cluster)

regards,

re: still having troubles

kevinhol — Thu, 16 Oct 2008 16:22:48 GMT

That override was not a fix to address all 2115's. It was only to address a specific situation with 2115 events of a Workflow Id : Microsoft.SystemCenter.DataWarehouse.CollectEventData.

1. Are ALL (at least 99%) your 2115 events coming from the above workflow? If they are - then apply this override - and bounce the healthservice on your affected management server (in a cluster - take offline and then back online) and you might consider clearing out the old healthservice cache.

2. If they are NOT all from DataWarehouse.CollectEventData, and are from random sources.... the next step is to see if they are all from a DataWarehouse workflow ID, or some are, some not. In either case, this is typically SQL database performance related. Bounce the healthservice and see if these comes back immediately, or if they take some time before you see them.

lots of ID 2115 and the RMS fails

Thomas — Fri, 13 Feb 2009 11:46:20 GMT

Thx Kevin for this great Blob.

I have been experiencing this error for a week now. After about 12 hours of working fine, the warnings 2115 start appearing. After some minutes, there are only 2115 errors in the Log and the RMS turns gray. The Workflow IDs include all possibilities, they seem to appear in a cyclic behavior to each other. The workaround didn't fix it, unfortunatelly. I suspect a SQL performance problem. Before migrating the whole system to a new, better server, I just wanted to make sure if this might be the Problem.

re: 2115s and grey RMS

kevinhol — Fri, 13 Feb 2009 19:23:59 GMT

So - here is what I look at with 2115's.

1. Look for a pattern - do the 2115's happen at a specific time or random? If a pattern - look for other jobs that might be running at that time, like a backup - or SQL DBA maintenance plans.

2. Look at the 2115's... do they come from a single datasource/workflow... or multiple? The workaround I posted only applies if they are ALL from the collectevent and data warehouse workflow.

3. Random 2115's with LOW times... (under 300 seconds) are normal... as long as we recover. If they have longer times associated with them... that is indicative of a SQL perf issue, or blocking on the DB. SQL perf is caused by keeping too much data in the DB, bad disk/memory I/O, not enough spindles for the DB, DB and logs not being in distinct volumes/spindle sets, poor SAN performance, too many agents in the management group, other jobs stepping on the DB, too many consoles open, etc....

re: Event ID 2115 A Bind Data Source in Management Group

Mike Ory — Thu, 14 May 2009 22:03:12 GMT

I just started getting these every 5 - 10 minutes.

All coming from Microsoft.SystemCenter.DataWarehouse.CollectEventData

They come back right away after bouncing the service. No MP's have been added/deleted in about 3 months.

The performance of the SQL server looks great The db and logs are on different volumes.

I've set the 3 overrides as described above. I've rebooted my SQL server and then the RMS.

Any other ideas?

re: just started

kevinhol — Thu, 14 May 2009 22:29:54 GMT

When did you put in these overrides?

If this just started - and the overrides have been in place for some time.... and SQL I/O performance is good... and this is ONLY coming from the warehouse collect event data source - then I would look for:

1. Blocking on the SQL server processes - check Activity monitor.... if performance counters look good on SQL - we can still have an insert problem if something is causing blocking.

2. Something is flooding events. Run the most common event query from my SQL query blog - and see if you can determine the source.

re: Event ID 2115 A Bind Data Source in Management Group

Mike Ory — Fri, 15 May 2009 00:23:59 GMT

Thanks Kevin.

Yes the overrides have been in place for a long time actually. I'm not seeing any blocking, but using your queries I found that a ton of events from our Exchange servers may be the problem.

I'm going to disable the event collection rule: "Execute: Test-ServiceHealth diagnostic cmdlet", which seems to be one of the major contributors, and see what happens.

And by the way, you say that you are not an expert on this subject? I have to repectfully disagree. ;)

re: careful

kevinhol — Fri, 15 May 2009 01:00:40 GMT

That is from the Exchange 2007 conversion MP - unfortunately - the event is created from script - if you turn off that workflow - you will also turn off the script. :-(

This is not a problem in the new native MP coming out with R2. That event will be a top consumer - but should not flood the database - it just will be at the top of the list.

re: Event ID 2115 A Bind Data Source in Management Group

Mike Ory — Tue, 19 May 2009 01:01:11 GMT

Turns out we DO have blocking.

This is causing the issue, but I'm not sure what can be done about it:

exec StandardDatasetMaintenance

re: exec StandardDatasetMaintenance

kevinhol — Wed, 20 May 2009 00:16:29 GMT

That is used by the warehouse for all maintenance jobs... including aggregation.

To reduce the impact by that job... you can focus on:

1. Increasing the disk I/O and server resources for the data warehouse database.

2. Reduce the amount of data going into the warehouse, and reduce the retention.

You really need to find out exactly what is causing the blocking when this runs... to determine the best course of action. A SQL DBA with SQL profiler in hand should be able to indentify the major causes...

How big is your warehouse? Agent count?

re: Event ID 2115 A Bind Data Source in Management Group

Dennis — Wed, 10 Jun 2009 11:02:42 GMT

Hey

In our environment we have the same problem but nearly all 2115 errors are coming from the workflow Microsoft.SystemCenter.DataWarehouse.CollectPerformanceData with times ranging from 1 minute till 5 minutes before the cache is written to the DW. I tried making overrides to the Performance Data Collector analog to the Event Data Collector although the values aren't exactly the same.

I've put these settings for now:

Maximum number of data items batches to process before maintenance: 50

Subsequent execution attempt timeout interval in seconds: 6

Maximum write operation execution attempt count: 10

This still generates a lot of 2115 errors together with constant alert of the type: Performance data collection process unable to write data to the Data Warehouse which get closed the next minutes.

Anything you can recommend?

re: Microsoft.SystemCenter.DataWarehouse.CollectPerformanceData

kevinhol — Wed, 10 Jun 2009 16:20:09 GMT

You should not make any overrides for this.

The overrides in this article handled a very specific issue with events, and it is NOT applicable to perf collection.

If you continuously have issues with perf insertion into the data warehouse - your warehouse is likely not performing well. Look for blocking, and for avg disk sec/write values.

re: Event ID 2115 A Bind Data Source in Management Group

Mike Ory — Wed, 10 Jun 2009 17:57:09 GMT

I've got a case open with Microsoft on this. I can tell you that one of the things they had me do (which didn't work for me, but may work for you) is to install this hotfix:

http://support.microsoft.com/kb/969130

They also had me run this query, which I think tells me where most of my DW writes are coming from?

select top 20 met.ManagedEntityTypeSystemName, count(*) from

ManagedEntityProperty mep

join ManagedEntity me on (mep.ManagedEntityRowId = me.ManagedEntityRowId)

join ManagedEntityType met on (me.ManagedEntityTypeRowId =

met.ManagedEntityTypeRowId) Where mep.FromDateTime > '2009-01-01'

group by met.ManagedEntityTypeSystemName having (count(*)) > 5 order by 2 desc

re: Why it didnt work

kevinhol — Wed, 10 Jun 2009 21:56:42 GMT

So - that hotfix - 969130 - simply allows dropping of old event tables. Their existence will not really impact event insertion into the DW - so that is why that didnt work. Also - that could only possibly affect the DW.CollectEventData 2115, and no others.

The MEP table query - dealed with discovery data. This can be a problem when management packs run discoveries that constantly update discovery data with properties that change frequently. If your only 2115 is from DW.CollectDiscoveryData - then a deeper analysis of discovered properties is in order.

The best queries I have seen for that are here:

http://nocentdocent.wordpress.com/2009/05/23/how-to-get-noisy-discovery-rules/

re: Event ID 2115 A Bind Data Source in Management Group

Mike Ory — Fri, 12 Jun 2009 17:15:50 GMT

Ok, that's good stuff. I ran the 'Discovered Objects in the last 4 hours' query and found that Microsoft.Windows.InternetInformationServices.2003.FTPSite has 36 changes.

I can tell you for sure that we have not added any FTP Sites in quite awhile...

re: FTP

kevinhol — Fri, 12 Jun 2009 18:20:21 GMT

The IIS MP has some frequent and noisy discoveries. Many run every hour. I like to modify the frequency of those to once per day.

Discovered Objects Query

Augusto — Tue, 07 Jul 2009 17:22:25 GMT

Great blog.

Just wondering where I can find the 'Discovered Objects in the last 4 hours' query.

Thanks.

re: 'Discovered Objects in the last 4 hours'

kevinhol — Tue, 07 Jul 2009 17:25:56 GMT

http://nocentdocent.wordpress.com/2009/05/23/how-to-get-noisy-discovery-rules/

re: Event ID 2115 A Bind Data Source in Management Group

John — Tue, 14 Jul 2009 13:30:28 GMT

A Bind Data Source in Management Group ABC has posted items to the workflow, but has not received a response in 122 seconds. This indicates a performance or functional problem with the workflow.

Workflow Id : Microsoft.SystemCenter.DataWarehouse.CollectEntityHealthStateChange

Instance : MS4.ABC.local

Instance Id : {43BE45BE-573D-AD34-B4333-3673F673BE32}

This come 4 times within a couple of minuts (first 61 sec, 122, 183 and then 245) - but do also come 20 times in a hour - do have a cluster DB with 64 GB RAM and a cluster RMS with 16 GB and 6 MS with 8 GB. There are (for now) 32 agents connected to one MS and even I move agents to other MS then the events come on the "new" MS. It shouldn´t be performance issue - Any idea?

re: Event ID 2115 - my solution

Curtis — Thu, 06 Aug 2009 06:39:15 GMT

I had the same symptoms (and pretty much all of them) as discussed above. My problem seemed to be that I had inserted an account into the 'Data Warehouse SQL Server Authentication Account' 'Run As Account' where I should have had a space.... as noted here:

http://www.eggheadcafe.com/conversation.aspx?messageid=30315729&threadid=30282292

This stopped the 2115 errors immediately.