Interpreting Jira-specific metrics for in-product diagnostics
Jira-specific metrics for JMX monitoring and in-product diagnostic are used in Jira Software Data Center and Jira Service Management Data Center. Learn more about how Jira-specific metrics are used and how to interpret them:
If you're searching for definitions of cross-product metrics for in-product diagnostics, check the article Interpreting cross-product metrics for in-product diagnostics.
We also have a separate article about Jira-specific infrastructure metrics.
Mail queue state metrics
Learn more about the following mail queue state metrics:
mail.queue.numItems
mail.queue.numItems shows the number of items in a mail queue. It's calculated as a snapshot of a mail queue size each minute. A constantly growing mail queue size might indicate an issue with sending or processing emails, thus leading to delays in the communication between Jira and users.
Available metrics: value, statistics
Warning markers  | Signs of healthiness  |
|---|---|
To check if a mail queue job is running frequently enough, compare it with the schedule of the mail queue service or with To compare the mail queue job with the service schedule, navigate to Administration > System > Services. These comparisons will help you find the reason why a mail queue isn’t being emptied. |
|
mail.queue.numErrors
mail.queue.numErrors shows the number of items in an error mail queue. It's calculated as a snapshot of an error mail queue size each minute.
When the Jira performance is healthy, an error mail queue doesn’t contain any items. At least one item in an error mail queue indicates that some issues are impacting the performance. Some of these issues can be resolved by resending items from an error mail queue. For example, when an instance loses the connection to an SMTP server.
Available metrics: value, statistics
| Warning markers | Signs of healthiness |
|---|---|
| An error mail queue contains one or more items. | An error mail queue contains zero items. |
A mail queue attempts to resend items up to 10 items. If the operation fails for the 11th time, the items will be put into an error mail queue.
From the error mail queue, you can remove items in one of the following ways:
Manually clear the whole error queue.
Manually resend all items from the error queue to a mail queue.
You should pay attention to the cases where an error mail queue item gets back to an error mail queue after you resend the items manually. These cases might indicate permanent performance issues.
mail.queue.numErrorsAddedPerMin
mail.queue.numErrorsAddedPerMin shows the average number of items added to an error mail queue per minute. Jira keeps track of each time an item is added to the error mail queue and calculates the average over the last minute.
When the Jira performance is healthy, an error mail queue doesn’t contain any items. At least one item in an error mail queue indicates that some issues are impacting the performance. Some of these issues can be resolved by resending items from an error mail queue. For example, when an instance loses the connection to an SMTP server.
Available metrics: value, statistics
| Warning markers | Signs of healthiness |
|---|---|
| An error mail queue contains one or more items. | Flatlining: the job wasn’t running or didn’t catch any mail queue items over a period of time. |
A mail queue attemts to resend items up to 10 items. If the operation fails for the 11th time, the items will be put into an error mail queue.
From the error mail queue, you can remove items in one of the following ways:
Manually clear the whole error queue.
Manually resend all items from the error queue to a mail queue.
You should pay attention to the cases where an error mail queue item gets back to an error mail queue after you resend the items manually. These cases might indicate permanent performance issues.
mail.queue.jobRunning
mail.queue.jobRunning shows the state of a mail queue job for each minute. It emits 0 when the job isn’t running or 1 when it is.
How it is working in details
Jira emits 1 when the mail queue job is starting processing (it’s happening after checking if there are any items in the mail queue, hence the value doesn’t represent each run of the mail queue job) and 0 at the very end of the job. Additionally, we have a different job which each minute asks the mail queue if its job is running and based on the response emits 0 or 1.
When the mail queue is empty, the metric will emit 0 for the run of a mail queue job. This means that to assess the performance and discover any issues in it, you should consider other metrics too.
Available metrics: value
| Warning markers | Signs of healthiness |
|---|---|
|
|
Mail queue insight metrics
Mail queue insight metrics indicate what’s happening inside a mail queue. Use them to find the causes of performance issues.
Learn more about the following mail queue insight metrics:
mail.queue.numItemsAddedPerMin
mail.queue.numItemsAddedPerMin shows the average number of items added to a mail queue per minute. Jira tracks each item added to the mail queue and calculates the average over the last minute.
Available metrics: value, statistics
The metric provides insight into what’s happening inside a mail queue but doesn’t indicate any warning signs. To discover performance issues, use this metric along with the others.
mail.queue.numItemsProcessedPerMin
mail.queue.numItemsProcessedPerMin shows the average number of items processed by a mail queue per minute. Jira tracks processed items and calculates the average over the last minute.
Available metrics: value, statistics
| Warning markers | Signs of healthiness |
|---|---|
To check how often a mail queue job is scheduled, navigate to Administration > System > Services > Mail queue service. For example, the job was scheduled to run each three hours, and then it was changed to run more often to speed up the processing of mail queue items over time. You can also check the schedule of a mail queue job by viewing Note that there might be a period where a mail queue remains empty, and | Items in a mail queue are processed right after they’re added. To check when the items are added and processed, compare |
mail.queue.numEmailsSentPerMin
mail.queue.numEmailsSentPerMin shows the average number of emails sent by the SMTP server per minute. Jira tracks each email the SMTP server sends and calculates the average over the last minute.
Available metrics: value, statistics
| Warning markers | Signs of healthiness |
|---|---|
| When multiple items are added to a mail queue and processed by it but no emails are sent, the SMTP server might have issues with the connection to a Jira instance. | Emails are sent after some items are added to a mail queue and processed by it. |