Menu Close

How to transition to the SQL Version Agnostic MP

image

Recently, Microsoft released a new SQL MP, Version 7.0.15.0.

https://www.microsoft.com/en-us/download/details.aspx?id=56203

This new MP supports discovery and monitoring of SQL 2012, 2014, 2016, 2017, 2019, and later versions of SQL Server.  It is completely different than all previously released SQL Management Packs, the older ones I will refer to as “Legacy” MP’s.  The new MP is much improved in that all DB Engines once again share a common base class, and you no longer have duplicated rules and monitors for each SQL version, with dozens of MP’s to keep up with and maintain.

This MP is the only supported MP for the above versions moving forward for monitoring the SQL DB engines.  If you were previously using the older MP’s for 2012-2016, there is a process you will need to plan for, in transitioning.  This article is a step by step plan on how to make the transition from the Legacy MP’s to the new Agnostic MP’s with the least duplication of alerting, and load on SCOM.

The first step, will be to import the 7.0.15.0 Management Packs.

image

If you had a previous version of this MP imported (Previously named SQL 2017+) then you might find that these MP’s will not import:

image

The MP Guide discusses this.  These are not updateable, and must be deleted first before importing the new version of this MP.

Once imported, you may see two “Microsoft SQL Server” folders in the console, due to the new and “old” MP’s sharing the same display name.  This will be corrected later in the process.

image

The DEFAULT settings for discovery for this MP is to discover ALL SQL versions SQL 2012 and later. If you do not want DULPICATE discovery of ALL your SQL computers, engines, DB’s, etc….   There is a rule in the management pack that attempts to inspect to see of the older 2012-2016 Legacy MP’s exist in the management group, and if they do, it will try to set an override on the primary DB Engine discovery to not discover 2012-2016, so no duplicates would be discovered.

There are two problems with this:

1.  The rule didn’t work in my testing, so it isn’t 100% reliable.

2.  The MP it tries to create is Microsoft.SQLServer.Overrides (Microsoft SQLServer overrides) and this might not fit in with your corporate override MP naming standards. 

For this reason, I recommend checking the discovery IMMEDIATELY after import, and setting the overrides you want in your own SQL Overrides MP.  The discovery Display Name is “MSSQL on Windows: Discover Local SQL Server Database Engines

image

First check to see if any overrides are set.  I recommend placing them in your own SQL Overrides MP that aligns to your override MP naming standards.

image

Now, at this stage, we should have the new MP fully implemented, and it should be discovering and monitoring on SQL 2017 and later.

The next step, is to onboard (discover and monitor) one SQL version at a time.  I will start with SQL 2016.  To make this work, I will delete SQL 2016 from the discovery override above, leaving SQL 2012 and 2014 in place:

image

Give this some time for the agents to get the new override, and you should start to see your SQL 2016 servers trickling into the SQL Server Roles view:

image

You need to verify that all your instances discovered as expected, and there are no monitoring configuration and security issues.  The new MP does a good job of alerting you if the security configuration is not sufficient.

Next, to avoid duplicate monitoring, you can delete your older SQL 2016 MP’s from the management group:

image

Note*** Deleting Management Packs is VERY expensive in the SCOM Database.  It is very common for MP’s to fail a delete attempt, or lock up the console for an extended period, if there is a very large amount of discovered instances and monitoring data from an imported MP.  If this is the case, you can always attempt to delete the MP’s, but if they will not delete, then try to delete them using PowerShell:

Get-SCOMManagementPack -DisplayName “Microsoft SQL Server 2016*” | Remove-SCOMManagementPack

Obviously you will need to fix/cleanup any unsealed MP’s which depend on (reference) the SQL 2016 MP’s first, such as removing overrides, custom views, etc.

If all of these fail, then the course of action would be to create a temporary override MP, and disable the SQL 2016 seed class discovery:

image

Once you disable this discovery, you can attempt to cleanup/delete SQL 2016 instances in SCOM using Remove-SCOMDisabledClassInstance.  Sometimes you must run this multiple times to remove all your Legacy SQL 2016 instances, as it may time out or fail with an error.  You can look in the legacy MP views to see if it is deleting them.  Once you have no more instances, removing the MP should be easy to complete.

If this goes well, you can move on to SQL 2014, then SQL 2012.  Repeat the removal of the override for that version, verify they get discovered, and then remove the legacy MP’s:

Get-SCOMManagementPack -DisplayName “Microsoft SQL Server 2014*” | Remove-SCOMManagementPack

Get-SCOMManagementPack -DisplayName “Microsoft SQL Server 2012*” | Remove-SCOMManagementPack

Lastly, if you are going to retain monitoring for SQL 2008 (which is not supported by this SQL Version Agnostic MP) then you will want to update to the most current version:

https://www.microsoft.com/en-us/download/details.aspx?id=10631

image

This will update the console views changing the Legacy MP’s to “old”:

image

Lastly – if you were using my SQL RunAs Helper MP, it has been updated for the SQL Version Agnostic MP, and is ready for deployment as well:

 https://kevinholman.com/2016/08/25/sql-mp-run-as-accounts-no-longer-required/

The security model has changed in the newest MP’s, so you might want to verify security is correct first.  See the above article for examples.

 

Some side notes:

  • SQL Mirroring is no longer supported, nor monitored in this SQL Version Agnostic MP.  For SQL 2012 and later, you should use AlwaysOn.
  • The Legacy SQL Replication MP’s are not supported with this MP for SQL 2012-2016.  The old Replication MP’s have dependencies on the Legacy SQL DB Engine MP’s and will need to be removed in order to remove all the old Legacy MP’s.
  • In general – NEVER import any of the below MP’s unless you determine their monitoring is critical to your business requirements.  Just because a MP exists does not mean you should import it.

SQL Replication

SQL Reporting Services

SQL Analysis Services

image

71 Comments

  1. EricGS

    Kevin, thank you for this guide. Excellent as usual. How well does the new MP cope with SQL 2008? We’re still in the process of migrating off this and it’s taking a while.

    • Techiedude

      Hi Ericgs – As SQL 2008 will EOL in less than four weeks (July 9th to be exact), the new agnostic version release does not support it. Please note the following in the download details: “This version-agnostic management pack was thoroughly built from the ground up in accordance with Best Practices for SQL Server; it is intended to monitor SQL Server 2012—2017 and all upcoming SQL Server versions.”

    • Kevin Holman

      Ericgs – I expect that customers will unfortunately run for a long time on SQL 2008, even understanding there will be no support for it soon, and this puts the environments at risk. This MP does not discover/monitor SQL 2008, so you can continue to use the old “legacy” MP’s for SQL 2008 to continue monitoring it, you will simply see it in the “old” views as I documented above already.

  2. Rick Bywalski

    We are using your Run AS MP will we need to rerun it on the existing servers that we have run it on with the old MP our should it just work. My guess is we will not need to run it and permissions already set will still work

    • Kevin Holman

      It will “work”, however, the new MP simplified the security configuration required to be the same across all versions. It is possible you might see the new monitor “Securables Configuration Status” showing an unhealthy state, and this will tell you if you are missing any recommended/required security rights. If this monitor shows unhealthy, you can review the list and determine if you are missing something.

      • Rick Bywalski

        Is the new run as back words compatible with the legacy mp? I was thinking if I get the new one installed and use it going forward I would have the proper rights in place when we can migrate to the new MP we would already have the proper permissions in place. We are only about half way through deploying to SQL. We do have one 3rd party software blocking us from deploying the upgraded MP for a few weeks.

  3. Chris Wood

    I noticed that the new SQL 2017+ MP changes the PrincipalName for DBEngines. It now shows the Netbios type instance name instead of the Parent FQDN of the host. Also the alerts no longer have PrincipalName populated. Is this intentional? Is there any way to get these attributes like they used to be?

    • Ian Blyth

      We have just imported this and found out this problem as well. We use Principal Name in the alert in Orchestrator to send alerts to ServiceNow. So we are have to look how to work around this. Interestingly the alerts from the Discovery MP does use Principal Name in the alerts it generates.

      • JAN NEVARIL

        I opened MS case and asked Microsoft if there is any solution for this, and there isn’t from their side.
        I solved the problem by a script that assigns server names to “custom field”, and in the orchestrator is the value assigned to the principal name parameter.
        I have most of the SCOM environments connected directly to the ServiceNow(via ServiceNow connector) and with the custom field, only new rule in the SN is needed.

        • Kevin Holman

          Hi Jan,

          How do you limit collection of alerts with the SNOW connector? Are you using groups, or do you just send all alerts to SNOW, then use SNOW event management rules to ticket/ignore?

          If you are using SCORCH to update the custom field, how do you ensure the custom field is updated with FQDN *before* the alert is sent to SNOW?

        • Ian Blyth

          We branched off based on the alert coming from the name of the Version Agnostic MP (Microsoft SQL Server on Windows (Monitoring)) and Microsoft SQL Server Integration Services on Windows MP as that has the same problem. We found that MonitoringObjectFullName has the FQDN in the middle of this field and so with Regex we were able to extract that and put it in the field where the PrinciplaName should be.

          @Kevin we only send Warning and critical alerts. If the alert is not needed then it is disabled or changed to Information if we want to keep an eye on it without creating ServiceNow tickets. We have a Runbook that checks that the names are FQDN.

  4. Rick Bywalski

    One word of caution if you are using any 3rd party add on for things like dash boarding check with your vendor if they support this pack or not before proceeding. I did and was glad I did as it would have broken some key functions in the product and they are busy updating their product to support the new build. This is a complete rewrite of the code and changes a lot of things like counters etc.

  5. Edwin

    Hey Kevin,
    Thanks for the guide helped me a bunch! However i used to have a daily scheduled report on a MSSQLSERVER object to report average wait time and cpu load to a specific group of people by e-mail.
    Now this (naturally) stopped working as i uninstalled the old versions of the SQL MP.
    However when i wanted to fix this i noticed i can’t find any mssqlserver objects or any rules that are in the new MP, i do see them under monitoring and i do see the performance rules there too and they show data. But when i want a detailed performance report (where i used to see them) i just can’t find them. Do you have any clue on what might be going on here? is this by design maybe, or is something wrong in my environment?

    • Edwin

      after removing the entire MP and reinstalling it, i now do see the new discovered objects in reporting aswell.
      now to see if the reporting works again tomorrow

  6. Edwin

    *next morning* all discovered objects that i saw yesterday are gone again in the reporting part… im so confused right now.

  7. richard scott

    does anyone have a copy of the new SQL Server RunAs Addendum Management Pack‎, the TechNet link seems to be broken ?

  8. Kiwifulla

    Probably an obvious answer, but we have created a lot of overrides against the SQL 2012, 2014, 2017 Management Packs over the years. These are against discoveries, monitors and rules. By following this article to remove those MP versions, I assume I have to recreate all of those overrides again against the equivalent discoveries, rules and monitors (if they are applicable) within the new agnostic Management Pack?
    Thanks in advance

  9. Rick Bywalski

    Is the new run as back words compatible with the legacy mp? I was thinking we could start using it now and have the proper rights there when we are able to deploy the new MP

    • Kevin Holman

      The new RunAs config SQL statement is backwards compatible with the 2012, 2014, and 2016 scripts in the old RunAs config – with one exception. The old script added:

      GRANT SELECT on sys.database_mirroring_witnesses to [SCOM_HealthService]

      Since the new SQL MP does not monitor database mirroring, this permission would be missing if you only used the new one. You could easily add this back into your custom copy.

  10. Gerald Versluis

    Hello Kevin,

    How about retaining historical data for capacity reports? What if we want to look at teh historical growth of a specific database over, say, the past six months? I assume that data will be lost…

    • Kevin Holman

      The data is not lost. We do not delete data from the DW when a managed entity is deleted. We simply sync a new managed entity to the DW and associate the new perfmon data with that. A single contiguous chart would be hard, but not impossible, depending on how it was queried.

  11. Michael Daffner

    Hi Kevin
    some SQL 2017 Servers will not be discovered in my environment (SCOM 2016 UR5) with the new agnostic SQL MP. The message I get in the eventlog
    “”Microsoft.SQLServer.Windows.Discovery.LocalDiscoverySeed”” running for instance “”xxx”” with id:””{1E89D4D7-46C2-D5D6-87E7-9D36EE6ECC75}”” cannot be initialized and will not be loaded. Management group “”ARZ”””,”System.String[]”

    Do you have any ideas for this problem?

  12. Paul

    Hi Kevin,

    To my surprise when reviewing my existing SQL Server 2016 MP(s) in prep for the new 2017+ MP I have found that they all have a dependency on the Default MP.

    During implementation of the SQL 2016 MP (on SCOM 2016) I was very careful to follow best practice and ensure that no overrides etc were saved to the Default MP so I am left wondering how this occurred.

    Your article implies that removing the ‘old’ SQL MP(s) would be straight forward (suggesting to me that there should not be any dependencies on the default MP under normal circumstances).

    Would you happen to know if this is a common situation (MS SQL MP’s becoming dependent on the default MP) and if so what is the most likely cause and the best way to fix this situation?

    Many Thanks,

    Paul

      • Paul

        Summary of References are:-
        >

        Microsoft.SQLServer.2016.Presentation

        Microsoft.SQLServer.Generic.Dashboards

        Microsoft.SQLServer.Visualization.Library

        Microsoft.SQLServer.2016.Discovery

        Microsoft.SQLServer.Generic.Presentation

        ….And I have found some sections in the default MP called “ComponentOverride” below which I don’t know what they are for exactly. I ran the overrides report against the default MP and there is nothing listed for SQL at all.

        >
        ComponentOverride ID=”Microsoft.SQLServer.2016.Presentation.UniversalDashboardOverride_cb9b0787_4668_4a2b_88ad_e0f894fe4c18″ TypeID=”SQLServer!Microsoft.SQLServer.2016.Presentation.UniversalDashboard”>

  13. Damian

    Hi Kevin,
    Fairly new to SCOM, but I’ve followed the steps and got this working on SCOM 2019, with just the new SQL agnostic pack installed, albeit with one issue.

    There is now only one SQL Server report listed
    Microsoft SQL Server on Windows (Views) -> Database Files Space Usage Forecast

    I was expecting to have a Microsoft SQL Server (Monitoring) folder with reports listed. Any obvious reasons what I’ve misconfigured?

    • David L

      Did you manage to resolve this? I can also only see the “Database Files Space Usage” report since deploying the SQL version agnostic MP

        • Kevin Holman

          It isnt a “problem” more than it is a change. The old SQL MP’s contains about 8 reports, but they were all linked reports so they can be recreated using the core foundational reports pretty easily. The new one only contains 1 report which is a new and custom report. I am not sure how much the old reports were even used. Do you have examples of specific reports you depended on?

          • Niksa

            Hi Kevin, not really dependent on any of the reports, was just wondering where they are gone.
            Thanks for clarification on this, I will recreate reports as I go.

  14. Raoul

    Hi Kevin, when I apply the PowerShell commands you provided, that will also delete SQL Reporting MP’s. Is this included in the new MP’s or not?

  15. Thom Inglin

    Hi Kevin,

    Thanks for this! I am sure it will be very helpful. My challenge right now is that there is a worrying (for us) note in the docs:

    “We recommend that you monitor no more than 50 databases and 150 database files per
    System Center Operations Manager agent to avoid spikes in CPU usage that may affect the
    performance of monitored servers.”

    We have some SQL clusters that have more than 100 databases on an instance. We have experienced some monitoring performance issues with the existing MP already so I am wondering if it is going to be more severe with this MP.

    • Kevin Holman

      That statement is not new. It’s been true since our very first SQL MP’s for SCOM 2007 back over 10 years ago.

      I have customers go well over the 50 DB mark…. I have even seen up to 300 DB’s and seem to work ok. However, I have also seen many DB’s get to the point that monitors fail to initialize, timeouts, resource consumption, etc. You just need to pay attention to those, monitor them closely, look at CPU/memory consumption for the SCOM processes, and look for errors in the SCOM Agent event logs. If you cannot support these large instances, you can take actions like disable DB file/logfile discovery, and try monitoring at the DB level, or instance level only, etc.

  16. Bill

    Since this new agnostic version was built from scratch did an ‘Express’ engine exclusion get built in, or can it monitor them properly?

    • Kevin Holman

      It does discover express, however you can exclude by name, additionally there is a dynamic group included that can be used to disable specific workflows if desired. I dont know of any errors specifically around express, as I almost always work to exclude them.

  17. Daniel Duchow

    Hi Kevin,
    great instruction on how to remove the “duplicated” SQL Views from SCOM. But unfortunately we did not get the instances for SQL 2012, 2014 or 2016 to be shown in the “SQL Server Roles” nor the “SQL DB Eingine on Windows ” from your “RunAs Config MP”.
    We did remove the exclusion from the override package like written in your post but after waiting for round about 12 hours we did not get more then 3 instances

  18. Steve Ward

    Hi Kevin
    Any details on the correct MS SQL Reporting service management pack to use with the new agnostic SQL mp versions, as I need monitor SQL 2012 RS.
    The only non version labelled MP I can find for MS SQL RS is 7.0.15.0 which actually is the 2017+Reporting services one once you import it, which you said should be installed.

  19. Andrew

    FYI – take care to note the differences in the Version Agnostic MP vs the legacy MPs. I just completed a multi-version SQL farm migration from SCOM 2016 with legacy MP’s to SCOM 2019 with only the Version Agnostic MP. I was happy to turn over to my DBAs a nice and “clean” SCOM view, only to then immediately be asked “where’s the ‘Tcp Port’ property for the DB Engines?”. Had I noticed this wasn’t available I might have chosen not to use this MP. Instead of configuring SCOM 2019 to also use the legacy MP, I am just accepting the DBA criticism and hoping it can be included in a future update…

    • Kevin Holman

      I believe we are planning on adding this back in a future update of the MP. This was removed, because it was no longer needed because of changes to the MP in how we connect to SQL. Since it wasn’t needed by the MP, it was removed. However, we understand customers may have liked that feature for visibility, grouping, and custom workflows, so this feedback has been taken.

  20. Scott Banyas

    Kevin,
    Might you be able to direct me on how to monitor what used to be called “Service Pack Compliance” in the old individual MPs and what now I believe is called “Product Version Compliance” in the new MP? Different versions of SQL have a different compliance version in our environment.

    • Kevin Holman

      The new monitor evaluates for each version separately – this is described in the Product knowledge.

      If you wish to override this – also covered in the product knowledge – you override for groups. You will have to create custom groups for each version of SQL instance. Such as an “SQL 2012 DB Engines Group”, etc.etc. and use the version property to create these groups.

  21. Benny

    Hi!

    Trying to go over to SQL Agnostic MP from the Legacy SQL MP´s. Have read this Guide for hours now, in case i´ve been missing something important. 1.I have removed all the overrides based on the Legacy SQL MP´s. 2. Uninstalled all the Legacy MP´s and the earlier Run as MP for SQL. 3. Installed the Agnostic SQL MP. 4. Installed the Run as SQL MP 7.15.1. I dont now why, but where missing many of the SQL Server Roles. Everything shows up under the SQL Run as MP under Seed. Under DB Engine over 50% SQL engines missing. Same in the SQL Server Roles there are missing 50 % of the SQL Server Roles. Any tip what to search for?

    • Kevin Holman

      If you are missing DB Engines, this means you have a discovery problem.

      1. The most common issue is that you do not have rights. You did not set up Service SID monitoring using the RunAs Helper MP correctly, or you did not grant rights to the service SID on the instance as needed. This will be evidenced by discovery failure alerts that I guarantee you – you have.

      2. The other common issue would be that the override is set to exclude specific versions of SQL, as described in the post and images above.

      It is going to be one, or the other, or both.

  22. Benny

    Ok number 2 is out beqause i can see different kind of SQL versions under the SQL view… But your right about the discovery problem, i can see thoose failure alerts you talked about. The thing is that for now the SCOM agents are in a multihome scenario.

    I have checked on a missing SQL Server Instance and the NT Service\HealthService is there and are mapped to the SA role. Can it be the cause that the service SID was created and granted with the older version of the the RunAs Helper MP from the old SCOM 2012 R2 environment? Maybe i have to delete SQL-logins and create them again from SCOM 2019 with the new RunAs Helper MP? But it is a little bit strange why some of SQL instances works and other not.

  23. Benny

    I can see now in the eventlog for one of the SQL Servers that there is a error with eventid: “4221” Source: “Error SQL Server Integration Services MP”

    There is a Exception with message: “SQL Namespace is empty or version less than minimum version”

    In the Inner Exception the message is: “Unable to execute query ‘Select * FROM ClientSettings’ for the path ‘\\SERVER.Domain.com\root\Microsoft\SQLServer\ComputerManagement13’ 32 bit: False. Error code -2147217392. Error message: Invalid class.

    I have no idea whats is wrong…?

    • Kevin Holman

      Invalid class means your SQL server is missing WMI entries that the management pack is expecting.

      Do you have the following WMI class present? root\Microsoft\SQLServer\ComputerManagement13

  24. Benny

    The same Error “4221” are consistent for every SQL Server that have Discovery problems. Looked on another SQL server and the WMI class “root\Microsoft\SQLServer\ComputerManagement13” are also present there, so i really dont know what to do…

    It worked very well with the Legacy SQL MP’s so this is strange. It also feels wrong to give up and go back to the Legacy MP´s.

  25. Martin

    Hello Benny!

    We had the same issue with some of or SQL Servers in the Discovery process.
    Try this one on the SQL Servers that have problem with the WMI questions.

    From an Elevated CMD prompt

    mofcomp “%programfiles(x86)%\Microsoft SQL Server\number\Shared\sqlmgmproviderxpsp2up.mof”

    Remember to change the number for your current SQL version.

  26. Stewart

    This version agnostic management pack (v7.0.15.0) appears to have issues discovering clustered SQL Server instances. I deployed it alongside a fresh SCOM 2019 build without any of the old SQL Server management packs imported, and only some instances/databases are being discovered – it’s a very strange mix of results. We have a SQL 2014 instance that gets discovered fully (including databases), a SQL 2016 instance which is discovered with its databases but are not monitored, and SQL 2016/2017 instances that aren’t discovered at all.

    I then imported the SQL 2016 management pack (v7.0.15.0) and it discovered all the clustered instances/databases with no issues whatsoever. I haven’t tested other versions yet but I’d expect it would likely be the same result.

    I see this error appearing in the Operations Manager log on the management servers (ID 10801):

    Discovery data couldn’t be inserted to the database. This could have happened because of one of the following reasons:

    – Discovery data is stale. The discovery data is generated by an MP recently deleted.
    – Database connectivity problems or database running out of space.
    – Discovery data received is not valid.

    The following details should help to further diagnose:

    DiscoveryId: 435a766d-decb-3d54-7d0d-08ca1142e6eb
    HealthServiceId: 330d8eeb-218d-4956-16df-e6a89aa7e061
    Microsoft.EnterpriseManagement.Common.DiscoveryDataInvalidRelationshipSourceException,The relationship source specified in the discovery data item is not valid.
    Relationship source ID: c1c83d48-b71f-296c-25be-80fd56d89cb2
    Rule ID: 435a766d-decb-3d54-7d0d-08ca1142e6eb
    Instance:
    5c324096-d928-76db-e9e7-e629dcc261b1HOSTNAME.domain.com8e36e48d-79a9-25f6-03f4-a83dbabfb7c5SQLVIRTUALNETWORKNAME.site.domain.com0ad613ed-b2ba-4bd3-05d2-102dc4d8e756INSTANCENAME.

    What’s interesting is that the 1st value contains only a partial domain name which doesn’t include the site, whereas the 2nd value contains the FQDN.

    I’ve ruled out cluster/SQL instance permissions, proxy agent settings, RunAs profiles, NIC settings, DNS, NetBIOS/default domain on the cluster nodes.

    Standalone SQL instances are discovered fine!

    • Marius Ene

      Same issue here. Seems that to monitor SQL cluster instances the SCOM management servers that are included in the SQL 2017 resource pool needs to have RPC remote access permissions to the SQL instances. This would be a big drawback I think.

  27. Kamil

    Hi All, I saw some issues after import New version mp for SQL. I’m using too other version MpS for SQL 2012-2016. After imported I disabled discovery in Agnostic MP for previous version SQL but now, from date import I See more events 2115 from data discovery. The Time delay is from 20000 to 40000 s. On database I See one task which is lock, and another tasks must waiting for this locking task… Anybody know where it will be problem? And how can I check and find root cause?

  28. Ian Blyth

    We have the Replication MP for 2016 and the database team want to monitor 2017 replication. When it says it needs the old replication MP does that mean it just needs the MP installed or does it need discovery running as well to get the objects?

  29. Andrew

    Kevin,

    Thanks for putting this together. Quick question: I have this new management pack, but I still need to monitor SSRS 2016. I added the “SQL Reporting Services 2016” MP. This results in a “Microsoft SQL Server (old)” folder being created. Aside from the unfortunate visual of an “old” folder, is there any other conflict with using “SQL Reporting Services 2016” MP alongside the v7.0.15.0 version agnostic SQL Server MP? Thanks!

  30. Olivier

    Hello Kevin,
    As some other people commented, i notice that the reporting part of the new SQL MP only has one report type, no monitoring report as before. Is it something that we do wrong while importing this MP? Are the reports available somewhere? I need some of these reports for SQL performance etc.
    Thx a lot for your help,

  31. Andre Swartz

    Hi Kevin
    The Management Pack appears to be discovery only stand-alone sql servers. None on my 2012, 2014 and 2016 clustered instances are being discovered. The 2017 instance where discovered but not monitored. Not sure where I should go from here any assistance would be appreciated.
    Thank you.

    • Kevin Holman

      I don’t repro that. I have clusters (traditional shared storage failover clusters and Always-On replicated clusters) and all discover without issue.

      Have you opened a Microsoft support case on this issue?

      Are you seeing any WMI errors in the Application event log? Are you seeing discovery errors in the agents OperationsManager event log?

  32. Gus

    Hi Mr. Master Kevin. Regarding Run As Addendum MPs and these new versions you said “It will “work”, however, the new MP simplified the security configuration required to be the same across all versions.” Do you recommend to use SID accounts or standard Run As accounts? Many thanks for your great support!

    • Kevin Holman

      SID. Always. RunAs accounts are good when they enable something that cannot be done otherwise, but they are a real pain to manage. Using Service SIDs is so much easier to maintain, and more secure.

  33. Steffen

    Hi Kevin, we are testing the new agnostic SQL mp in our scom dev at the moment.
    First we deleted all our old SQL mps and waited for some minutes (no more SQL views and profiles were present).
    After that we imported the “new” SQL mp 7.0.15.0 and placed our SQL admin (sa and local admin) in the discovery and monitor run-as profile.
    After some minutes only the SQL 2016 on our scom reporting server was discovered.
    We are missing three more testservers with SQL 2016 and 2017 but none is showing up.

    any idea why ? no error or warning in scom.
    Those missing SQL Servers worked with the old mp without a problem.

    • Kevin Holman

      First off, I am not a fan of using RunAs accounts for these, especially not ones with SQL Admin rights. I have much better solutions posted, using lowest priv and service SID’s.

      Second – if you have rights, did you distribute these RunAs accounts to the agents? Did you inspect the event log on the agents to look for discovery attempts or errors? Did you restart the agents to speed up the discovery process? Did you check to see if you discovered the SEED class on the expected servers?

Leave a Reply

Your email address will not be published. Required fields are marked *