Audit the data your instance sends
Your Data Center instance collects diagnostics data and sends it to Atlassian Cloud to generate Instance health insights. You can check this data yourself in the Optimization tab of Troubleshooting and support tools, without leaving your product. The tab has two data auditing tools:
Download collected data gives you a zip file of JSON files with diagnostics data that your instance has collected for its next submission, such as performance, health, and security metrics. You can use it before and after you connect your instance. Downloading doesn't send this data to Atlassian.
Recent submissions lists what your instance sent in the last 3 days and what each submission contained. It's available after you connect.
These tools cover the data used for Instance health insights. They don't cover cloud readiness assessments or the Customizations report. To find out what we collect overall, see Security and data handling.
はじめる前に
You must have system administrator permissions for your Data Center instance. You might be asked to confirm your password again.
Make sure the Atlassian Troubleshooting and Support Tools app is up to date.
Preview the data before you connect
Until you connect your instance, the Optimization tab shows the setup steps. You can preview the data before you decide how to send it.
To preview the data:
Open Troubleshooting and support tools:
In Jira, go to Administration, select System, then select Troubleshooting and support tools.
In Confluence, go to Administration, select General configuration, then select Troubleshooting and support tools.
Select the Optimization tab.
Under Before you start, find Preview the data, then select Download collected data (JSON).
Wait while your instance prepares the file. On a cluster, it collects the data from every node, which can take a few minutes.
Save the zip file when the download starts, then unzip it to review the files.
Your instance only assembles and saves the data. It doesn't send this data to Atlassian.
Before you connect, the download includes only data that doesn't need your consent. After you connect, it also includes data that your data sharing level allows, such as the list of installed apps and system events.
If you plan to upload data manually, the file you upload is a support.zip, not this download. See Collect and upload Instance health data manually.
Download the collected data after you connect
After you connect, the Optimization tab shows your connection status and the Data activity section instead of the setup steps.
To download the collected data:
In the Optimization tab, find Data activity.
Select Download collected data (JSON).
Save the zip file when the download starts, then unzip it.
The download shows what your instance would send next: data it has collected since its last submission (up to 24 hours by default), plus a snapshot of current configuration and health data taken when you select the button. To see what your instance already sent, use Recent submissions.
Review recent submissions
Below Data activity, the Recent submissions section lists what your instance sent in the last three days, newest first. Each row is one submission that reached Atlassian Cloud.
On a cluster, the table lists the submissions of all nodes.
Select Refresh table to load the latest submissions. The table shows 10 submissions per page.
The table shows how much was sent, not the data itself. To see the kind of data your instance sends, download the collected data.
Only successful submissions are listed. If a submission fails, your instance doesn't add a row.
Manual uploads aren't listed. You upload a support.zip from your browser to Atlassian Administration, so your instance doesn't record it.
Records stay on your instance. Your instance stores them in its database and, by default, deletes them after 30 days. The table shows the last three days.
列 | What it shows |
|---|---|
Sent at | When your instance sent the submission, in your time zone |
タイプ | Diagnostics for performance, health, and other diagnostics data, or Support zip for a limited support zip that your instance creates and sends automatically |
ソース | What started the submission, for example Automatic for diagnostics data sent as part of the regular transfer, or Scheduled for a support zip that your instance creates on a schedule |
要約 | For diagnostics, the number of samples. A sample is one metric entry in the data. For a support zip, the file size |
Select Show details to see what a submission contained.
Diagnostics submissions
For a diagnostics submission, the details show the number of samples in each category. Each metric entry in the downloaded files has a category tag in capital letters, such as RESPONSE_TIMES. In-product diagnostics uses IPD.
カテゴリー | What it contains |
|---|---|
In-product diagnostics | Metrics from In-Product Diagnostics (IPD) monitoring, such as database connection and cluster node latency |
System health | Failed health checks, system events such as reindexes, startups, and Assets imports, selected configuration such as feature flags, the list of installed Marketplace apps, and, in Jira, counts such as issues, comments, and custom fields |
App usage | How Marketplace apps are used, and aggregated navigation usage. In Jira, with Extended data sharing, this also includes custom apps |
応答時間 | Ready-for-User (RFU) timings, Apdex scores, and, in Jira, slow JQL query statistics |
セキュリティ | Pass, fail, or error results of security checks, with details such as counts and reasons |
Support diagnostics | The number, size, and age of log, Java Flight Recorder, and export files, and the disk space available in the home and temporary directories. File names and paths aren't included |
パフォーマンス | JMX performance metrics (Jira, Confluence) such as CPU load, garbage collection time, and memory use, performance metrics for individual Marketplace apps, and aggregated browser, device, and network characteristics |
Assets | Performance metrics for Assets, such as slow queries (Jira only) |
ガードレール | Counts of entities that can affect performance, such as projects, users, groups, custom fields, spaces, and index size. Find out more information about guardrails |
Support zip submissions
For a support zip submission, the details show whether the zip was Anonymized, and the bundles it included with the size of each.
Support zips that your instance sends automatically are created with data redaction on, so Anonymized normally shows Yes. When a zip is anonymized, data such as usernames, email addresses, and IP addresses is replaced with pseudonymized values. Redaction is pattern-based and might not catch every value. Find out more about redacting data in a support zip.
The Bundle column shows the technical name of each bundle. A bundle is listed only if it added data to the zip. For more about what a minimal support zip contains, see Minimal support zip.
Bundle | What it contains |
|---|---|
| The main application log ( |
| Thread dumps and per-thread CPU usage taken from Java Flight Recorder recordings. The recordings themselves aren't included. Find out more about the Java Flight Recorder |
| The Tomcat |
| Tomcat access logs. Only with Extended data sharing |
| Thread dumps: snapshots of JVM thread activity. Find out more about thread diagnostics |
| For each Atlassian or Marketplace app, the names of the Java classes it contains. Custom apps aren't included |
| A small file with basic instance details: the app version, the product name and version, the license type, the time zone, the base URL, the Support Entitlement Number (SEN) and server ID, and, on a cluster, node details such as node IDs and addresses |
What the downloaded files contain
The download is a zip file named collected-diagnostics-data-<ID>.zip. It contains:
One JSON file for each node. A non-clustered instance has one file.
manifest.json, which lists when the file was generated, whether it's complete, and which nodes are included, missing, or failed. On a cluster, a node that doesn't respond within about 5 minutes is left out and listed as missing.
Each node file contains:
Instance and node information: the product name and version, base URL, Support Entitlement Number (SEN), a hashed node identifier, and some configuration context, such as the app version and monitoring settings. To match a hashed node identifier to a node, use the Node names in Atlassian Administration table at the bottom of the Optimization tab.
Metric entries. Each metric entry includes:
Metric name: The specific metric being tracked.
Attributes: Key-value pairs representing the metric's data, such as performance indicator values.
Type: The kind of metric, such as
JMX_METRICfor performance, response time, and in-product diagnostics metrics,GUARDRAILSfor counts,APP_USAGEfor app usage information,SYSTEM_EVENTfor system events and failed health checks,SYSTEM_METRICfor configuration and system statistics,SECURITY_METRICfor security check results, orAPDEXfor Apdex scores.Timestamp: The time at which the metric was recorded.
Tags: Additional information about the metric, such as its
categoryand, for metrics about an app, the app key (for examplepluginKey).
While it prepares the download, your instance saves the per-node files temporarily in its export directory and deletes them automatically after about an hour. Your downloaded copy stays with you.
関連情報
Security and data handling: what we collect, store, and transmit
この内容はお役に立ちましたか?