How to use the Alerts tile
The Alerts tile allows you to view
You can add alerts tiles to any dashboard or perspective.
You may also like to watch the Alert Management Webinar.
How to configure an Alerts tile
Add a new tile to a dashboard and choose the Alerts tile.
Select the visualization for your Alerts tile and click next.
ListThe Alerts list shows a list of alerts.
DonutThe Donut tile shows a summary of the number of alerts either by severity,
priority or resolution state . Clicking on a donut segment or on the key shows a list of alerts with that severity,priority or resolution state .Scope:
The scope section allows you to define the objects for which alerts are shown.
Tip: The advanced > criteria option is a useful way to narrow down alerts. For example, you can filter the list shown by owner or alert name:
Owner LIKE 'sales\user1'
orName LIKE '%failed&'
For more information about using advanced criteria such see How to use criteria when scoping alerts.
Scope options on dashboardsTip: If you experience any problems with scoping tiles, you'll find FAQs and help in the article How to scope tiles.
List List allows you to select one or more objects or groups.
You can add multiple
objects and groups . To remove anobject or group click the x to the right of its name.Tip: Start typing and after two characters you'll see suggestions that match the name appear.
Tips for using wildcards when searching for objects or groupsBy default searching will look for the top 10 items containing the words listed in the search. If you wish to create a more specific search you will need to use wildcards (*).
If you place a wildcard after the term you are looking for, it will find all the objects which start with that word searched and any terms that may follow. If you place a wildcard at the start of the search term, it will look for objects that contain the searched word and also have terms before that word.
If you enclose your searched term in wildcards it will look for objects which contain the searched word, this object will not begin or end with the term searched.
Group Group allows you to select members of a specific group. Only one group can be selected.
Advanced Advanced allows you to select a group, class or both. You must at least define either a group or class. You can define both. You can also use criteria to narrow down your selection.
Group:
Same as the group option above.
Class:
Class equates to the target class within SCOM. As you type the dropdown will be populated with suggestions of matching classes from SCOM, from which you can select the required class.
Criteria:
Criteria allows you to create an expression to further refine the scope.
Examples for commonly used criteriaObjects you would like to see Criteria Objects with particular text in their name DisplayName like '%Server1%'
Objects starting with a particular string DisplayName like 'test%'
All objects in maintenance mode InMaintenanceMode = 'TRUE'
Only healthy objects HealthState = 1
Objects with a health state in SCOM of 0, an unknown health state (uninitialized), a gray health state icon with a question mark. HealthState = 0
Objects that are not healthy HealthState != 1
Objects in critical state HealthState = 3
Objects in critical or warning state HealthState = 2 or HealthState = 3
To show all gray uninitialised objects HealthState = 0 OR HealthState IS NULL
All objects not in maintenance mode InMaintenanceMode != 'TRUE'
Objects where the parent agent is offline IsAvailable='false'
Objects that are offline, in maintenance or state unknown IsAvailable='false' OR InMaintenanceMode=1 OR HealthState=0
Computers with a particular OS OSVersion = '6.3.9600'
List objects by name and filter by HealthState (Name like '%Server3%' OR Name like '%Server4%' OR Name like '%Server2%') AND HealthState=3
List objects by SCOM Id and filter by HealthState Id IN ('7021174b-9e5d-5fbf-878a-42b9f0bf6f4a', '9bd4a1cc-f07a-0e36-b37d-d9ee974e0f3c') AND HealthState=3
Exclude object from the Group specified DisplayName not like '%server3%'
Exclude objects from the Group specified (DisplayName NOT LIKE '%server3%') AND (DisplayName NOT LIKE '%server4%')
For more information see:
Scope options on perspectivesNote: If you never used a perspective, you should read Working with perspectives before scoping tiles on perspectives.
The power of perspectives is that tiles on a perspective can use a dynamic scope. A dynamic scope considers the currently viewed
object . A dynamic scope consists of two different states:the configuration of the scope in the tile (for example, "consider child objects of type logical disk for the currently viewed object")
the actual resolved scope that depends on which
object you are currently viewing ("this object has 5 child objects of type logical disk")
After configuring the dynamic scope once in the tile, you'll get different results depending how the scope is resolved on the different
objects you are viewing.This object
The dynamic scope will be the resolved to the object that is currently viewed.
Relative objects
The dynamic scope will be resolved to relative objects (parents or children) of the object that is currently viewed.
How to pick relative objectsAt the top, you'll see the name of the object you are currently looking at. Now you can choose if you want to pick parents or children of that object, and if this parent or child relation should be considered only one level up or down the SCOM model or through all levels.
Class:
Here you pick the class of the objects you want to select. If you leave this field empty, the scope falls back to the "this object" scope.
Note: You will only see groups and classes that the object you are currently looking at is a member of.
Tip: If you want to pick objects of any class, enter the SCOM base class logical entity in the class field.
Tip: If you want to find out what classes the object you are interested in belongs to, you can go to the Monitored Entity perspective of that object. You'll see all the classes the object belongs to listed there.
Criteria:
You can narrow the selection of objects of a particular class down further by entering criteria for those objects. For more help see How to use criteria when scoping objects.
Tip: If you want to find out what properties you can base your criteria on, you can go to the Monitored Entity perspective of the object you are interested in. You'll see all the properties for criteria listed there.
.
Note for SCOM task tiles:
You have to pick one single object for SCOM task tiles since the task has to be run for one specific object. If you pick the scope option relative objects, you need to narrow the scope down to one object with the help of a class and criteria.
Other specific objects
Gives you the normal, non-dynamic scope options you are used to when scoping tiles on dashboards. This means the tile will not dynamically adapt it's content to the currently viewed
object , it will always show data for the staticobject picked here.Since the power of perspectives is that their tiles can show data for different
objects depending on whatobject is currently being viewed, you should only select this option when you are sure that there is no relationship between the desired scope and the currently viewedobject .Filters:
The filter section allows you to control which alerts are shown by the tile.
You can select from severity, priority, source, owner and state:
Scope options on perspectivesNote: If you never used a perspective, you should read Working with perspectives before scoping tiles on perspectives.
The power of perspectives is that tiles on a perspective can use a dynamic scope. A dynamic scope considers the currently viewed
object . A dynamic scope consists of two different states:the configuration of the scope in the tile (for example, "consider child objects of type logical disk for the currently viewed object")
the actual resolved scope that depends on which
object you are currently viewing ("this object has 5 child objects of type logical disk")
After configuring the dynamic scope once in the tile, you'll get different results depending how the scope is resolved on the different
objects you are viewing.This object
The dynamic scope will be the resolved to the object that is currently viewed.
Relative objects
The dynamic scope will be resolved to relative objects (parents or children) of the object that is currently viewed.
How to pick relative objectsAt the top, you'll see the name of the object you are currently looking at. Now you can choose if you want to pick parents or children of that object, and if this parent or child relation should be considered only one level up or down the SCOM model or through all levels.
Class:
Here you pick the class of the objects you want to select. If you leave this field empty, the scope falls back to the "this object" scope.
Note: You will only see groups and classes that the object you are currently looking at is a member of.
Tip: If you want to pick objects of any class, enter the SCOM base class logical entity in the class field.
Tip: If you want to find out what classes the object you are interested in belongs to, you can go to the Monitored Entity perspective of that object. You'll see all the classes the object belongs to listed there.
Criteria:
You can narrow the selection of objects of a particular class down further by entering criteria for those objects. For more help see How to use criteria when scoping objects.
Tip: If you want to find out what properties you can base your criteria on, you can go to the Monitored Entity perspective of the object you are interested in. You'll see all the properties for criteria listed there.
.
Note for SCOM task tiles:
You have to pick one single object for SCOM task tiles since the task has to be run for one specific object. If you pick the scope option relative objects, you need to narrow the scope down to one object with the help of a class and criteria.
Other specific objects
Gives you the normal, non-dynamic scope options you are used to when scoping tiles on dashboards. This means the tile will not dynamically adapt it's content to the currently viewed
object , it will always show data for the staticobject picked here.Since the power of perspectives is that their tiles can show data for different
objects depending on whatobject is currently being viewed, you should only select this option when you are sure that there is no relationship between the desired scope and the currently viewedobject .Timeframe:
The timeframe section allows you to determine the timeframe for the data. You can choose either to use page timeframe, which allows users viewing the page to change the timeframe on the fly, or for you to set a fixed specific timeframe which will ignore the page timeframe settings.
What is the page timeframe?The page timeframe is the timeframe setting a dashboard
or perspective is currently using. When a user changes the page timeframe, all tiles that use the page timeframe will adapt to the new timeframe. Tiles that don't use the dynamic page timeframe aren't affected and won't change.Timeframe shows alerts which were created in this time period. If you change it to 24 hours hoping to see alerts closed in the last day, you will miss alerts which were created longer ago, even if they were closed in the last day.
Settings for different visualizations:
Settings for ListsSort
Sort allows you to change how the list is sorted. You can sort ascending or descending by:
name
raised
severity
path
repeat count
Limit
Allows you to define a maximum of alerts that will be shown by the list tile.
Columns
The columns section allows you to show the column titles, rename titles, sort and hide columns in the Alert list.
Note: If you want to be able to resize the width of columns, you need to show the column titles (by activating the check box show column titles). This enables you to control the default width of the columns by dragging them to the size you want. It also enables your users to temporarily change the column width by dragging the columns to the size they want while they are looking at the dashboard.
Note: When resizing the width of columns, make sure to leave enough space in the last column for the update state functionality to be shown:
Last column ok:
Last column too small:
Note: Even if you show column titles, the severity column won't show a title because it only contain the severity icons.
Display
Use the slider to adjust the font size.
Settings for DonutsGrouping
When configuring the Alerts Donut tile the grouping section allows you to select whether the donut shows by severity,
priority or resolution state .Display
Size mode:
Default Displays the donut scaled to the height of the tile. Fill Enlarges the donut to use the whole width of the tile. If you chose the fill option and show the legend, you can define the size of the legend with a slider. Show legend:
Allows you to show or hide the legend of the graph.
Click done to save the tile.
The tile now shows alerts according to your settings.
Rule alerts vs monitor alerts
Users may not be aware of the differences between monitor and rule alerts, and that monitor alerts should not be closed. Dashboard Server uses different alert icons for rule and monitor alerts to help users be more aware of these differences.
Monitor alerts
Monitor alerts are shown by a solid icon (like the health state icon, but round) to indicate that the alert is affecting the health state of the object, because the alert has been raised by a monitor.
A monitor alert is 'stateful' and therefore current. A monitor changes the state of the monitor, which then triggers an alert. This means that if the monitor alert is still visible, then it is still a current issue. Once action has been taken, a user can click the reset monitor button to force SCOM to reevaluate the health of the object, and if the issue is re-detected then the alert will be reopened. When the issue is resolved the alert will be closed automatically, so monitor alerts should not be closed manually.
When changing the resolution stateclosed is not the default option, and if a user tries to close a monitor alert a warning is shown and confirmation is required.
Monitor alerts are used to show Health state summaries for unhealthy objects in Dashboard Server.
Rule alerts
A rule can trigger an alert, but not change the state of the object, or rollup to the parent. A rule alert shows that something has happened in the past, but it won't necessarily happen again. If the issue reoccurs, then the alert RepeatCount increases. Rule alerts should be closed manually when the issue has been resolved.
Rule alerts are shown by a hollow icon (that’s different from the health state icon) the alert is not affecting the health state of the object because it has been raised by a rule. Rule alerts should be closed manually by clicking the resolution state button.
Summary of the differences between rule and monitor alerts
Monitor alerts | Rule alerts |
---|---|
Stateful | Non-stateful |
‘There is a problem’ | ‘There was a problem’ |
Affects health state of object | Does not affect health state of object |
Rollup health to parent object | Does not rollup health to parent object |
Changes health state of dashboard | Does not change health state of dashboard |
Creates alert (normally) | Creates alert (normally) |
Alert stays open while issue remains | RepeatCount incremented if issue reoccurs |
Closes alert automatically (normally) | Does not close alert automatically |
Can be manually reset back to healthy | Cannot be reset (no state) |
Alerts should not be closed manually | Alerts should be closed manually |
Health state summaries
Objects and their health state
At the heart of SCOM monitoring is its object model and the health state monitoring of those objects; if a disk is low on space, the disk is marked as critical and this rolls up to the server, which is also marked as critical.
What is a health state summary?
In Dashboard Server health state summaries are shown for unhealthy (yellow or red) objects. Health state summaries show why an object is yellow or red and show you a summary in-line with the status icon itself. This means that, at a glance, you can see the cause of the critical health state and spot common issues across multiple objects.
If you find several servers are showing red, then the health state summaries can help answer some of your questions:
Why is it red?
Is it the same reason that the others are red?
Is it related to the application issue I'm seeing?
Which red server is the priority?
Do I have to click on each server to see what the problem is?
Where are Health state summaries available?
Health state summaries are shown wherever an objects health state is shown, for example the Matrix tile, VADA in view and analyze mode, the Alerts tile, and in the Status tile.
Health state summaries are not available for container objects such as groups and distributed applications (DAs), however they work excellently for objects that host things (not contain things) such as servers, devices, software, and their sub-components.
In some cases, you might find that a critical server does not have a health state summary.
How do health state summaries work?
Health state summaries work by performing a lookup for monitor alerts (alerts that are affecting the health state) for each object.
For any object that is not healthy, Dashboard Servershows the alert that is:
- Created by a monitor (not a rule – they don’t affect health)
- Most severe (if it’s critical, it will look for a critical alert)
- Most recent
What if no health state summary is shown?
- Some monitors do not create an alert
- Perhaps the system has closed the alert and it has been groomed out of the Operations Manager database due to a retention setting. In SCOM, under Administration > Settings > Alerts the Automatic Alert Resolution tab shows you how many days after the last modified time (repeat count) from which all active alerts will be resolved, and how many days after the object is healthy that the alert will be resolved.
- Maybe a user closed the alert? The walkthrough below shows how you can show 'recently closed alerts to help troubleshoot this.
Walkthroughs
This walkthrough shows how to add a list of only error alerts where the owner is the currently logged on user.
- On the dashboard or perspective where you want the list to appear click on the orange plus at the bottom of the screen to add a new tile.
Click on the Alerts tile button.
Click on the List button.
- Give the tile a suitable title, for example
My error alerts
. - Leave the scope as it is and click next to move to the filters section.
- Change the severity to only show error alerts, by clicking on warning and info to deselect them.
- In the owner section click on current user and any will be deselected.
- Click done and the list of your error alerts are shown.
This walkthrough shows how to add a list of recently closed alerts to the default Alerts dashboard on the Overview menu, but you could also add this to a new perspective for particular alerts.
- Browse to Overview menu and click on the Alerts dashboard.
- Click on the orange plus at the bottom of the screen to add a new tile.
- Click on the Alerts tile button.
- Click on the List button.
- Give the tile the title
Closed monitor alerts
. - Leave the scope as it is and click next.
- In the filters section change the source to only show monitor alerts, by clicking on rule to deselect it.
Under state click on closed so that only closed alerts are shown.
At this point the list of the most recent 5 closed monitor alerts will be shown, and you can click done if you wish. For this walkthrough we will continue with the configuration.
- Click next.
- Leave timeframe set to last 30 days, and click next. timeframe shows alerts which were created in this time period. If you change it to 24 hours hoping to see alerts closed in the last day, you will miss alerts which were created longer ago, even if they were closed in the last day.
- In the limit section change the limit to
15
, and click next. - In the columns section we're going to add in a column to show who closed the alert. Click show hidden columns.
- Click show next to ModifiedBy.
You may like to reorder the columns by dragging the gray column handle.
If the ModifiedBy column shows
system
, then this is mostly likely to be when the object changed state to become healthy, and the alert was automatically closed. If you wish you can click done at this point.If you would like to filter out alerts that were last modified by
system
you can use advanced criteria as described below.Return to the scope section of the tile and click on advanced.
In the criteria box paste in
LastModifiedBy != 'system'
meaning that you do not want alerts that were last modified bysystem
.For more information about using scope > advanced > criteria see How to use criteria when scoping alerts.
To recreate a monitor alert which has been wrongly closed you can reset the unhealthy monitor. This will set the monitor back to healthy, then when it is next reevaluated, if the problem still exists then the monitor will be changed to unhealthy and the alert will be recreated.
Click on the object that is unhealthy, but not showing a health state summary.
- Click on the monitored entity perspective.
- Scroll down to the children section and look for the child that is unhealthy (red).
- Click on that child, then the monitored entity perspective.
- Scroll down to the Monitors section and click on the red monitor.
Click on the reset button up the top right.
Click on the reset button.
Frequently Asked Questions
Can I list only particular alerts?
You can use criteria in the scope > advanced section to list alerts by name or with particular words in the name, for example Name = 'Failed to Connect to Computer'
see How to use criteria when scoping alerts.
Can I improve the format of the SCOM Alert emails?
You can enable HTML email notifications in SCOM that link to the Alert in Dashboard Server see How to enable HTML email notifications in SCOM.
Can I link an alert to a ticket in our ticketing system?
Yes, you can add a hyperlink to the Alert in Dashboard Server that links to a particular ticket see How to add links to an external ticketing system from Alerts.
Can I list alerts that have not been linked to a ticket?
You can use criteria in the scope > advanced section to list where the TicketId field is blank, for example TicketId is NULL
see How to use criteria when scoping alerts.