Pings
Sourcegraph periodically sends a ping to pings.sourcegraph.com
to help our product and customer teams. It sends only the high-level data below. It never sends code, repository names, usernames, or any other specific data. To learn more, go to the Site admin > Pings page on your instance (the URL is https://sourcegraph.example.com/site-admin/pings
).
Sourcegraph will also periodically perform a license verification check, to verify the validity of the configured Sourcegraph license. Tampering with these checks, or preventing them from occuring, will cause Sourcegraph to disable many features until a successful check is completed. Certain Enterprise licenses can request to be exempt from these license verification checks.
Telemetry
Sourcegraph aggregates usage and performance metrics for some product features in our enterprise deployments. No personal or specific information is ever included.
List of other telemetry
-
Randomly generated site identifier
-
The email address of the initial site installer (or if deleted, the first active site admin), to know who to contact regarding sales, product updates, security updates, and policy updates
-
The external URL of the instance (e.g. "https://sourcegraph.example.com")
-
The IP address of the instance (e.g. "172.xx.xx.xx")
-
Sourcegraph version string (e.g. "vX.X.X")
-
Dependency versions (e.g. "6.0.9" for Redis, or "13.0" for Postgres)
-
Deployment type (single Docker image, Docker Compose, Kubernetes cluster, Helm, or pure Docker cluster)
-
License key associated with your Sourcegraph Enterprise subscription
-
Aggregate count of current monthly users
-
Total count of existing user accounts
-
Aggregated repository statistics
- Total size of git repositories stored in bytes
- Total number of lines of code stored in text search index
-
Whether the instance is deployed on localhost (true/false)
-
Which category of authentication provider is in use (built-in, OpenID Connect, an HTTP proxy, SAML, GitHub, GitLab)
-
Which code hosts are in use (GitHub, Bitbucket Server / Bitbucket Data Center, GitLab, Phabricator, Gitolite, AWS CodeCommit, Other)
- Which versions of the code hosts are used
-
Whether new user signup is allowed (true/false)
-
Whether a repository has ever been added (true/false)
-
Whether a code search has ever been executed (true/false)
-
Whether code navigation has ever been used (true/false)
-
Aggregate counts of current daily, weekly, and monthly users
-
Aggregate counts of current daily, weekly, and monthly users, by whether they are using code host integrations
-
Aggregate daily, weekly, and monthly latencies (in ms) of search queries
-
Aggregate daily, weekly, and monthly integer counts of the following query syntax:
- The number of boolean operators (
and
,or
,not
keywords) - The number of built-in predicate keywords (
contains
,contains.file
,contains.repo
,contains.commit.after
,dependencies
) - The number of
select
keywords by kind (repo
,file
,content
,symbol
,commit.diff.added
,commit.diff.removed
) - The number of queries using the
context:
filter without the defaultglobal
value - The number of queries with only patterns (e.g., without filters like
repo:
orfile:
) - The number of queries with three or more patterns
- The number of boolean operators (
-
Aggregate daily, weekly, and monthly user counts of search queries with the above properties
-
Code navigation usage data
- Total number of repositories with and without an uploaded precise code navigation index
- Total number of code navigation queries (e.g., hover tooltips) per week grouped by language
- Number of users performing code navigation queries (e.g., hover tooltips) per week grouped by language
-
Batch Changes usage data
- Total count of page views on the batch change apply page
- Total count of page views on the batch change details page after creating a batch change
- Total count of page views on the batch change details page after updating a batch change
- Total count of created changeset specs
- Total count of created batch change specs
- Total count of created batch changes
- Total count of closed batch changes
- Total count of changesets created by batch changes
- Aggregate counts of lines changed, added, deleted in all changesets
- Total count of changesets created by batch changes that have been merged
- Aggregate counts of lines changed, added, deleted in all merged changesets
- Total count of changesets manually added to a batch change
- Total count of changesets manually added to a batch change that have been merged
- Aggregate counts of unique monthly users, by:
- Whether they have contributed to batch changes
- Whether they only viewed batch changes
- Whether they have performed a bulk operation
- Weekly batch change (open, closed) and changesets counts (imported, published, unpublished, open, draft, merged, closed) for batch change cohorts created in the last 12 months
- Weekly bulk operations count (grouped by operation)
- Total count of executors connected
- Cumulative executor runtime monthly
- Total count of
publish
bulk operation - Total count of bulk operations (grouped by operation type)
- Changeset distribution for batch change (grouped by batch change source:
local
orexecutor
) - Total count of users that ran a job on an executor monthly
- Total count of published changesets and batch changes created via:
- executor
- local (using
src-cli
)
-
Aggregated counts of users created, deleted, retained, resurrected and churned within the month
-
Aggregated counts of access requests pending, approved, rejected
-
Saved searches usage data
- Count of saved searches
- Count of users using saved searches
- Count of notifications triggered
- Count of notifications clicked
- Count of saved search views
-
Homepage panel engagement
- Percentage of panel clicks (out of total views)
- Total count of unique users engaging with the panels
-
Weekly retention rates for user cohorts created in the last 12 weeks
-
Search onboarding engagement
- Total number of views of the onboarding tour
- Total number of views of each step in the onboarding tour
- Total number of tours closed
-
Sourcegraph extension activation statistics
- Total number of users that use a given non-default Sourcegraph extension
- Average number of activations for users that use a given non-default Sourcegraph extension
- Total number of users that use non-default Sourcegraph extensions
- Average number of non-default extensions enabled for users that use non-default Sourcegraph extensions
-
Code insights usage data
- Total count of insights
- Weekly count of page views on the insights pages
- Weekly count of unique viewers on the insights pages
- Weekly counts of hovers and clicks insights by type (e.g. search, code stats)
- Weekly counts of resizing of insights by type
- Weekly counts of edits, additions, and removals of insights by type
- Total count of clicks on the "Add more insights" and "Configure insights" buttons on the insights page
- Weekly count of users that have created an insight
- Weekly count of users that have created their first insight this week
- Weekly count of total and unique views to the different
Create
,Create search insight
, andCreate language insight
pages - Weekly count of total and unique clicks of the different
Create search insight
andCreate language usage insight
, andExplore the extensions
buttons on theCreate new insight
page - Weekly count of total and unique clicks of the
Create
andCancel
buttons on theCreate search insight
andCreate language insight
pages - Total count of insights grouped by time interval (step size) in days
- Total count of insights that are organization visible grouped by insight type
- Total count of insights grouped by presentation type, series type, and presentation-series type.
- Weekly count of unique users that have viewed code insights in-product landing page
- Weekly count of per user changes that have been made over the query field insight example.
- Weekly count of per user changes that have been made over the repositories field insight example.
- Weekly count of clicks on the "Create your first insight" CTA button on the in-product landing page.
- Weekly count of clicks on the code insights in-product template section's tabs, with tab title data.
- Weekly count of clicks on the use/explore template card's button.
- Weekly count of clicks on the "view more" template section button.
- Weekly count of clicks on the in-product landing page documentation links.
- Weekly count of filters usage on the standalone insight page
- Weekly count of navigation to dashboards from the standalone insight page
- Weekly count of clicks on "Edit" from the standalone insight page
- Total count of individual view series, grouped by presentation type and generation method
- Total count of insight series, grouped by generation method
- Total count of views, grouped by presentation type
- Total count of organisations with at least one dashboard
- Total count of dashboards
- Total count of insights per dashboard
- Weekly count of time to complete an insight series backfill in seconds
- Weekly count of requests of exports of Code Insights data
-
Search aggregations usage data
- Weekly count of hovers over the search aggregations information icon
- Weekly count of open/collapse clicks on the sidebar and expanded view of search aggregations
- Weekly count of search aggregation mode clicks and hovers
- Weekly count of search aggregation bars clicks and hovers
- Weekly count of search aggregation success and timeouts
-
Code monitoring usage data
- Total number of views of the code monitoring page
- Total number of views of the create code monitor page
- Total number of views of the create code monitor page with a pre-populated trigger query
- Total number of views of the create code monitor page without a pre-populated trigger query
- Total number of views of the manage code monitor page
- Total number of clicks on the code monitor email search link
- Total number of clicks on example monitors
- Total number of views of the getting started page
- Total number of submissions of the create code monitor form
- Total number of submissions of the manage code monitor form
- Total number of deletions from the manage code monitor form
- Total number of views of the logs page
- Current number of Slack, webhook, and email actions enabled
- Current number of unique users with Slack, webhook, and email actions enabled
- Total number of Slack, webhook, and email actions triggered
- Total number of Slack, webhook, and email action triggers that errored
- Total number of unique users that have had Slack, webhook, and email actions triggered
- Total number of search executions
- Total number of search executions that errored
- 50th and 90th percentile runtimes for search executions
-
Notebooks usage data
- Total number of views of the notebook page
- Total number of views of the notebooks list page
- Total number of views of the embedded notebook page
- Total number of created notebooks
- Total number of added notebook stars
- Total number of added notebook markdown blocks
- Total number of added notebook query blocks
- Total number of added notebook file blocks
- Total number of added notebook symbol blocks
- Total number of added notebook compute blocks
-
Code Host integration usage data (Browser extension / Native Integration)
- Aggregate counts of current daily, weekly, and monthly unique users and total events
- Aggregate counts of current daily, weekly, and monthly unique users and total events who visited Sourcegraph instance from browser extension
-
IDE extensions usage data
- Aggregate counts of current daily, weekly, and monthly searches performed:
- Count of unique users who performed searches
- Count of total searches performed
- Aggregate counts of current daily user state:
- Count of users who installed the extension
- Count of users who uninstalled the extension
- Aggregate count of current daily redirects from extension to Sourcegraph instance
- Aggregate counts of current daily, weekly, and monthly searches performed:
-
Migrated extensions usage data
- Aggregate data of:
- Count interactions with the Git blame feature
- Count of unique users who interacted with the Git blame feature
- Count interactions with the open in editor feature
- Count of unique users who interacted with the open in editor feature
- Count interactions with the search exports feature
- Count of unique users who interacted with the search exports feature
- Count interactions with the go imports search query transformation feature
- Count of unique users who interacted with the go imports search query transformation feature
- Aggregate data of:
-
Code ownership usage data
- Number and ratio of repositories for which ownership data is available via CODEOWNERS file or the API.
- Number of owners assigned through Own.
- Narrowing search results by owner using
file:has.owners
predicate. - Selecting owner search result through
select:file.owners
. - Displaying ownership panel in file view.
- Narrowing search results by owner using
-
Histogram of cloned repository sizes
-
Aggregate daily, weekly, monthly repository metadata usage statistics
-
Cody providers data
- Completions
- Provider name
- Chat, fast chat, and completion model names (only for "sourcegraph" provider)
- Embeddings
- Provider name
- Model name (only for "sourcegraph" provider)
- Completions
-
Whether Cody context filters are configured in the site config (true/false)
Allowlist IPs / CIDR Ranges for Sourcegraph
Starting 5.2.0:
- For
pings.sourcegraph.com
, allowlist the IP address:34.36.231.254
- For
telemetry-gateway.sourcegraph.com
, allowlist the IP address:34.107.157.31
- For
sourcegraph.com
, allowlist the full Cloudflare IP ranges
Prior to 5.2.0, allowlist the full Cloudflare IP ranges.
Using an HTTP proxy for telemetry requests
The environment variable TELEMETRY_HTTP_PROXY
can be set on the sourcegraph-frontend
service, to use an HTTP proxy for telemetry requests.
Be sure to update the enviornment variable like so : TELEMETRY_HTTP_PROXY:"http://proxy.example.com:8080"
Connections to Sourcegraph-managed services
Sourcegraph only connects to Sourcegraph-managed services for three purposes:
- The pings described above are sent, in order to:
- Check for new product updates.
- Send anonymous, non-specific, aggregate metrics back to Sourcegraph.com (see the full list above).
- Verify the validity of the configured Sourcegraph license.
- Legacy Sourcegraph extensions are fetched from Sourcegraph.com`s extension registry.
There are no other automatic external connections to Sourcegraph.com (or any other site on the internet).
Connections to Sourcegraph.com via Cody app
The Cody app connects to Sourcegraph.com to send a limited selection of the pings described above in order to infer value to our users and send update notifications. They include:
- Randomly generated site identifier
- Deployment type (the Cody app)
- Release version
- Operating system
- Total number of repositories added
- Whether a user was active today (boolean)
Troubleshooting Pings
It may happen that Sourcegraph will stop sending critical telemetry to Sourcegraph.com, if this happens it may indicate a problem with Sourcegraphs frontend database, or a site settings misconfiguration. Below are some debugging steps.
Sourcegraph telemetry pings are handled by a goroutine running on Sourcegraphs frontend service called updatecheck
, updatecheck
is started on container startup and periodically requests a variety of queries be run in the pgsql
database.
Misconfigured update.channel
The most common scenario in which Sourcegraph stops sending pings is a change to the update.channel
setting in an instance's site config
SHELL"update.channel": "release",
This setting must be set to "release" in order for the telemetry goroutine to run.
Check if the goroutine is running
This section is under development and currently only applies for docker-compose instances: 39710
If it's reported that pings aren't being sent to the Sourcegraph.com, you can check that the goroutine is running with the following command:
SHELLdocker exec -it sourcegraph-frontend-0 sh -c 'wget -nv -O- 'http://127.0.0.1:6060/debug/pprof/goroutine?debug=1' | grep updatecheck'
Example:
SHELL[ec2-user@latveria-ip ~]$ docker exec -it sourcegraph-frontend-0 sh -c 'wget -nv -O- 'http://127.0.0.1:6060/debug/pprof/goroutine?debug=1' | grep updatecheck' 2022-04-05 20:53:32 URL:http://127.0.0.1:6060/debug/pprof/goroutine?debug=1 [14660] -> "-" [1] # 0x1f052c5 github.com/sourcegraph/sourcegraph/cmd/frontend/internal/app/updatecheck.Start+0xc5 github.com/sourcegraph/sourcegraph/cmd/frontend/internal/app/updatecheck/client.go:697
Looking for errors in the logs
If the update.check
is running, and the site config is correctly configured, then it may be the case that pgsql
is failing to return data from the SQL queries to the frontend
. Check out the frontend logs for logs tagged telemetry: updatecheck failed
.
If issues persist, please reach out to a team member for support at [email protected]