Group checks allow the nesting of multiple checks that may or may not comprise a single system, and can also be useful when systems are interdependent.
To skip to a specific section of the "Intro to Group Checks" video, click the links with the !
Grouping checks allows you to receive alerts under customizable failure conditions, or to report on the group’s average uptime %.
Table of Contents
- What Can a Group Check Do?
-
Creating a Group Check
- Managing Group Checks
- Accessing Group Check Reports
- Use Cases for Group Checks
What Can a Group Check Do?
- Organize multiple checks monitoring a single system into one group
- Alert when a specific group of checks sees a widespread outage
- Escalate group check incidents to a system owner
- Track components of a system, such as load balance
Creating a Group Check
To create your first group check, click on Monitoring then Add New and select Group from the Check type drop-down.
It is also possible to create a group check or add to an existing one using bulk actions. Select several checks from your list of checks and then click More Actions > Create Group Check or Add to Group Check.
You will find the Group Check nested beneath the Web subheading in the All Checks dropdown.
Setup and Check Selection
There are two methods to select checks:
- Manually select checks
- Auto-select checks by tag
Manual check selection allows you to add up to 200 individual checks to a group from a list of your existing checks.
Auto-selecting a check by tag will add all checks that have the selected tag(s) applied, including present and future checks added to this tag.
A single Group Check is allowed a maximum of 50 tags, but each tag can contain an unlimited number of checks. You can find more detail about check tags here.
A single check of any type can exist in multiple groups, but group checks cannot contain other group checks.
Group Behavior
A group check can alert you when Downtime/Alert Conditions are met, or it can report an average uptime % for all checks within a group.
When the Report on a group's average uptime % option is selected, the Group Check’s report will display the percentage of uptime between all checks included in the group check, for the selected date range. Each check has its own uptime percentage, which is then aggregated between all checks included in the group check to calculate an overall uptime percentage.
In this example, the Group Check’s current status is UP, but the Uptime is only 50% since some individual checks experienced downtime. The Uptime calculation is based on the overall uptime of all checks within the group for the selected date range:
The aggregated uptime for each group check for the last 24 hours is also displayed when viewing Group Checks in the list of all checks on the Monitoring page:
As an additional example, Group Check 1 includes two checks: Check A and Check B.
- Check A has an uptime of 75% in the selected date range.
- Check B has an uptime of 25% in the selected date range.
- Therefore, Group Check 1 has an overall uptime of 50% for the selected date range.
Note: Since the group check uptime percentage is calculated based on the uptime of each check in the group, any check which is paused will not affect the overall group check percentage. Additionally, if a check is paused or in a maintenance window for part of a day within the selected date range, only the time that the check was actively being monitored will be included in the calculation for aggregate uptime for the check.
When you report on average uptime percentage, the check’s contact field becomes disabled, and the check does not issue alerts. Additionally, a crossed-out bell icon in the UI signals notifications are not enabled.
Additional note on check types: Real User Monitoring (RUM), Blacklist, Malware, SSL Certificate Expiry, and WHOIS/Domain Expiry checks are excluded from calculations for average uptime percentage reporting.
The group check’s default behavior is to alert you based on Downtime/Alert Conditions.
Group Check Alert Conditions
Group checks calculate a state change once every 30 seconds, and the state of the group (either UP or DOWN) is reflected at that time.
Group Checks do not include real-time analysis. Alerts will specify when Group Down / Alert Conditions are met, and will contain a brief listing of checks nested within the group that reported downtime.
Group Down / Alert Conditions
Define a condition that will bring the group down. The following options are available:
- Any assigned check is down
- 2 / 3 / 4 / 5 / 10 assigned checks are down
- 1 / 3 / 5 / 10 / 25 / 50 % of assigned checks are down
- All assigned checks are down
An alert is issued when the selected Group Down / Alert Conditions are met, and the check is set to Alert when Downtime/Alert Conditions are met.
This group check alert is in addition to any alerts issued by the individual checks within the group.
Response Time Calculation
Group checks can display response time metrics by copying the response times of a single source check, or by averaging the response times of all checks of a single check type that are within the Group Check itself. In the “Optional” settings tab of the Edit Check screen, select the appropriate calculation mode, and then the corresponding option for “single check” or “check type”. By default, the calculation mode is “No Response Time”.
When a Response Time Calculation mode is selected, the response time graph will be displayed anywhere metrics are displayed, such as Dashboard check cards, Status Page components or system metrics, the main Check screen (Monitoring), as well as the Group Check’s Check Report.
Please note: Once a mode is selected, it may take a few minutes for the Group Check to start collecting the response time metrics from the corresponding calculation source. Once recorded, historical data is retained in a single graph, even if calculation mode settings are changed.
Copy From Single Assigned Check
Assign a single check as a source for the Group Check’s response time graph. Any check in the Uptime.com account with measured response time metrics can be selected in this field, so assign this check with caution.
Please note: If a check is assigned as the response time source for the Group Check and is later deleted or removed from the Group, the response time graph will not collect new metric data until the single check source is updated. For more information, see "Deleting Checks in a Group Check" note in the Check States document.
Average by Single Check Type
Select a single check type to average the response times of all corresponding checks within the Group Check. The available check types for this calculation mode are HTTP(s), Transaction, API, ICMP(Ping), Heartbeat, or Incoming Webhook.
In this mode, averages are calculated by location(s) for each check, and then individual check averages are averaged together for the displayed response time graph. For more detailed information on how these averages are calculated, see Check Field Explanations.
Managing Group Checks
A group check can contain multiple checks of any type, except for other group checks. Clicking any group check will display all checks nested within it.
Managing group checks does not affect the settings of individual checks nested within that group, and there are a few additional limitations.
For example, group checks do not have an overall real-time analysis, as this is performed based on each individual check within the group.
Group Checks Page
The Group Checks page shows all group checks configured for your account. Click into a given check to see a list of all assigned checks organized within that group check.
Adding and Removing Checks from a Group
To add or remove checks from an existing group, locate the group check from Monitoring > Group Checks, and select the Edit check Icon> Edit. Select checks from the Checks Available window, and deselect any checks from the Checks Selected window to remove them:
Note: Deleting a check from within a group deletes the check from your account but deleting a group check does not delete the individual checks within it.
If checks have been assigned to a group by tag, and you wish to remove them, you will need to navigate to the check(s) in question and edit it to remove the check’s tag. This operation can be done in bulk.
Please note: If a deleted or removed check was the single source of the Group Check’s response time graph, then the graph will not record metrics until a new check is assigned. See the note in Deleting Checks in a Group Check for more information.
Maintenance with Group Checks
Individual checks assigned to a group can be in maintenance mode, but the group check does not inherit this status from the nested children checks.
To put a group check in maintenance mode, click Monitoring, select the check in question and click More actions > Set Maintenance.
Click the Maintenance tab to set a maintenance window.
Scenario:
If an individual check under maintenance belongs to a group that is not under maintenance, and this check goes down, its downtime applies to the failure condition you have selected.
Accessing Group Check Reports
To access an Uptime report for a Group Check, click More actions > View Check Analysis.
When viewing the group check, you can also click the Group Check’s name to view its Check Analysis.
Group checks do not list a response time, but response time is available for any checks nested within a group.
Use Cases for Group Checks
Get Alerted to Widespread Outages
With large enough sites or monitored infrastructures, a single failed URL doesn’t always indicate a problem. If you’re monitoring dozens, or hundreds, of URLs, it may be more beneficial to think of your site’s uptime in terms of percentage of down URLs. Using the Uptime.com group check, you can monitor these URLs and set alerts when 1%, 3%, or even 10% of these URLs go down.
Cataloging all Checks for a Single System
Suppose “My System” contains 5 components, and more than one of those components can fail before the system will stop working. A group check set to fail if 2 of its 5 assigned checks fail will give your team first-alert capability that a critical system is potentially down.
Escalating to a System Owner
Building on the previous use case, Uptime.com normally recommends escalations to a system owner after X minutes of downtime have passed. Such practices lead to resolutions, and improve overall downtime response. Group checks can offer a more immediate and precise escalation to the system’s owner, with data on which systems are down to improve reaction and time to resolution.
Properly configured, alerts from a Group Check can provide real-time visibility on potentially critical system failures.
Track Important Product Pages
Group checks can monitor performance and uptime for a set of HTTP(S) checks aimed at your most trafficked landing pages. Group these checks together for reporting purposes to see how well these most important pages perform. Group Checks do not report on response time, but the individual checks nested within them do. Use Real User Monitoring to supercharge the data you collect with real-user session performance.
Load Balance Tracking
Group checks can be effective for monitoring a load balancer. Setup will require multiple HTTP(S) checks, one pointing directly to the domain and the remainder pointing to each load balance IP address using the HOST header. More details on check setup, with examples, are available here.
Group these HTTP(S) checks into a single group check and set failure conditions to either multiple URLs, or by percentage of URLs down for the most accurate representation into the state of your load balancing.
Reporting on Average Uptime% for a Single System
Set a Group’s behavior to Report on average uptime % to create a check that will provide the average of any checks assigned to it as a single value. Attach this check to any existing SLA and scheduled reporting, as a card in your Dashboard, or as a component of your status page.
Want to see our checks in action? Check out our YouTube Library for more!
Comments
0 comments