Monthly Archives: February 2020

Management Pack Recap – January & February 2020 Wave

This is a summary of the wave of Management Packs that were released in January and February 2020. Information and download location in the links provided:

Certificate MP v1.4.3.0here
SQL Management Pack v7.0.20.0here
Oracle Management pack v1.3.1683.0here

If you know of any other Management Packs that have been released recently that I may have missed leave me a note in the comments and I’ll add them

Loading

Azure Monitor and SCOM: Together in a hybrid world

I recently published an article on LinkedIn with my thoughts on monitoring Hybrid cloud with Azure Monitor and SCOM, I am cross posting on my blog for my record but if you prefer you can also read it here https://lnkd.in/d8PvWsH

With the shift of the IT landscape towards public cloud and then toward hybrid there has been some confusion as to which monitoring solutions are the best fit. People with System Center Operations Manager (SCOM) are wondering about the product life-cycle and others are wondering where Azure Monitor comes into the picture.

Hopefully I can shed a little light onto this puzzle.

No alt text provided for this image

First lets look at SCOM, well good news, it’s not going anywhere for quite some time. The current iteration 2019 has a commitment of 10 years of support from Microsoft with an Update Rollup release every 6 months. If you have any on-premise real estate then SCOM is still a strong contender to manage those systems with it’s powerful proven track record, extensible custom monitoring and data-center focused features. If you have extended your network into the cloud you can also leverage these features for your data center-like IaaS components such as VMs.

No alt text provided for this image

Now for your cloud needs, let’s talk Azure Monitor. You may be thinking “but I’m not using Azure” and that’s ok because have I got more good news for you, Azure Monitor can integrate with VM’s running on other clouds, additionally with Azure Arc you can now make use of Azure services and features on any cloud.

No alt text provided for this image

Azure Monitor is a cloud first monitoring tool as such it’s designed to manage and monitor a wide range of cloud components and services leveraging not only the real-time capabilities of Azure Metrics but also the power of Log Analytics both of which feed into a wide variety of insights, visuals and analysis tools. This allows you to have a complete in-depth picture of your cloud real estate from a single location. Azure Monitor is also designed to be configured to monitor only the items you deem most critical which, when done properly results in a lean noise free monitoring experience.

No alt text provided for this image

So what should I use to monitor my Hybrid Cloud scenario?

The short answer is: No matter where you are you are on your cloud journey there’s a monitoring scenario that has you covered.

Regardless of whether you are migrating your VMs to an IaaS platform; refactoring your on-premise applications to leverage PaaS/SaaS services or have some applications which reach back into your on-premise environment, you have options and you don’t have to use a single tool either

Lets unpack this a bit.

If you have a large on-premises environment then certainly a data center designed tool such as SCOM will meet your needs and it can also cater for any VMs that you have migrated into the cloud. This will of course give you the on-prem monitoring experience that you are familiar with, which does offer a level of comfort while preparing to move to more cloud focused services.

Similarly if you are nearing the end of your cloud journey then Azure Monitor has got you covered from Billing to Subscriptions and everything in-between. Not only will you be able to proactively handle your tenant and resources through automation but you will also have access to rich interactive reports in the form of workbooks as well as machine learning capabilities.

However if you are between these two scenarios you can use both tools, in fact I’d even encourage it, they work very well together and leverage each others strengths to give a comprehensive single pane of glass across any Hybrid environment.

No alt text provided for this image

Loading

Silect Dashboards for SCOM

Silect has released an awesome set of dashboards for SCOM and best of all they’re free. You can request your copy here.

All you need to get started is PowerBI Desktop available here, then load the pbix file and edit the data source to point at your SCOM database instances.

Then hit refresh and you’ll have 10 nice health dashboards with an overview of your SCOM environment.

I really like this solution, it’s elegant and easy to setup. Great job Silect.

Loading

SCOM 2019: UR1 has been released!

The much anticipated release of UR1 for SCOM 2019 is here and you can get it from the update catalog.

There’s quite a bit to unpack here so lets have a look

The big changes are of course support for gMSA which I for one have been looking forward to for a while. The security benefits are obvious and I can say while it does work well be aware that it is complex to implement so read those instructions thoroughly.

The one click upgrade experience is also a step in the right direction making things much simpler for the SCOM admin, just note it only includes the server update so you will still need to upgrade your console, reporting etc manually.

See below for a complete list of all the changes

Improvements and issues that are fixed


Web Console Fixes/Changes:

  • State widget now supports sorting by health and age. 
  • Alert widgets can now be searched on Date Time and sorted based upon age and severity.
  • The alert link in email notification when browsed throws an error, “Your session with the Web Console server expired” even though the user was not logged in web console . You will now see the login page post this fix.
  • Alert summary view in SCOM Web console was fixed in size.  This can be expanded as required now.
  • When Alert state is changed to some custom state, then these alerts are not displayed in web console. Alerts with custom resolution states will be displayed now.
  • Few additional scrollbars appear when a customer widget is created in web console or browser window size is reduced.
  • Improvement: SCOM views will load/save much faster than before. 

Unix/Linux/Network Monitoring Fixes/Changes:

  • SCX by default does Info level logging which causes SCX logs to be filled up soon. With 2019 UR1, Warnings and Errors will be logged by default not Info.
  • In the network device monitoring, if the node property is changed to null/empty, it displays the old value of the property not the null value. The actual value will be displayed now which includes null.
  • In a scenario where multiple (~500) virtual network interfaces are running on agent, SCOM sometimes received heartbeat failure alerts for such agents. SCOM will monitor only physical interfaces hosted on these machines. For monitoring virtual interfaces, you need to set “enumvif” to true. 
  • SNMP GET or SNMP WALK resulted monitoring host to crash when the network object ID is equal to ULONG Max. There is an error with event ID 4000 due to this crash.
  • For linux distros servers, when the kernel version >= 4.18, then file system related performance data is shown as 0 in SCOM. This is fixed in FS provider to make it kernel version agnostic and collect file system stat info.
  • Use of Async Windows Management Infrastructure (MI) APIs is default functionality from SCOM 2019 UR1 for scalability improvements.More details here

 Admin Console Fixes:

  • When admin created “one-time maintenance mode schedule” for non-English locales, SCOM console displays an error “The client has been disconnected from the server. Please call ManagementGroup.Reconnect() to reestablish the connection”. Admin would be able to create one-time maintenance schedule now.
  • When SCOM agent is multihomed, then SCOM 2019 MS always displays logon type as “Interactive” for this agent. The correct logon type will be displayed now.
  • Unit monitor to check correct logon type for Run as Account had incorrect name and Operational State.
  • “Operations Manager Products” view in Admin console did not update the “Version” column for the installed component version. This column will now reflect the updated version of all the components listed.

Others:

  • When system proxy is configured on the Reporting Server; proxy overrides are ignored by the “Report Console Watcher Monitor” and “Web Console Watcher Monitor” and as a result HTTP/HTTPS access fails.  With UR1 fix, “Set proxy Direct” parameter of the mentioned monitors can be set to true and URLs will be accessed.
  • SCOM did not discover Windows Cluster which are deployed on servers with IPV6 network only enabled. IPV6 network will be supported for cluster monitoring now.
  • When the Domain controller is moved to a different AD site, then SCOM does not display the correct site name. Any dynamic group which are created for SiteName will also fail

 In addition to these, all the issues fixed in SCOM 2016 UR8 and prior UR for SCOM 2016 are also fixed in SCOM 2019 UR1. Details of the fixes are below.

  • In a scenario where SCOM monitors 100s of virtual machines hosted on a single Hyper-v server; every hour the healthservice.exe of each Virtual machine write into the VM page file simultaneously. Due to this concurrent paging, every hour disk I/O increases and database becomes unresponsive. HealthService.exe now have Memory Trimming enabled by default on an hourly schedule. A registry key is provided to disable the memory trimming and control the duration.

          Registry key is: “HKLM\Software\Microsoft\Microsoft Operations Manager\3.0\Setup\MemoryTrimming”

         Enable – 0 (Trimming is disabled); 1 (trimming is enabled)

         DelayInSeconds – Time period agent waits to start trimming (default is 120s)

         PeriodInSeconds – Recurring period at which the working set should be trimmed (default is 3600s)

  • Historical data do not appear, if input reporting end time is before group creation time. With this fix, historic data for a group (if data is available for objects in the group) would be displayed irrespective of group creation time.
  • Maintenance mode state changes which are recorded in MaintenanceModeStage table requires grooming when table grows. If the table is large, grooming takes longer and the operation times out with SQLTimeOut exception.
  • If a group is renamed in a Management pack, then console shows the new value but Powershell command Get-SCOMGroup returns the old name of group. Database Updates functionality was inconsistent for SCOM group renaming through MP and SCOM Console.
  • CPU Spike issues because of workflows running on all agents at the same time is addressed through script optimization and removing the sync time. 
  • Improvement: Sometimes SQL stored procedure “p_SelectForNewTypeCache” takes long time to complete, and SDK service fails to start. This is fixed and above SQL stored procedure will complete faster now.
  • Improved the performance of SCOM console in listing the groups.
  • Users of a scoped group are not able to use the Console.
  • SCOM console crashes while trying to connect to Azure Log Analytics and Azure Monitor.
  • SCOM Network Device Re-Discovery now probes for SNMP V3 devices too.
  • Agents by Health State report shows duplicate agent names.
  • Fixed an issue that prevented addition of a group in the Storage Spaces Direct 2016 management pack dashboard.
  • Linux agent is not able to get the correct version and port details for JBoss EAP 7.1.
  • An issue that lead to creation of multiple empty temp files in the /tmp directory of Linux servers has been fixed.
  • Fixed the formatting issue with the output for the task ‘Top10 CPU Processes’ when using Windows Management Infrastructure (MI) APIs.
  • Fixed an issue that caused the corruption of /etc/login.cfg file on AIX 7 machines during install/upgrade of the agent.
  • AIX Agent is now transitioned to 64-bit package to accommodate more stack and heap space if needed to avoid any stack/heap overflow which occasionally leads to heartbeat failure.
  • Free memory calculation accommodated appropriately on RHEL-7 platform.

Loading