WatchDog Docs
WatchDog documentation

Using WatchDog

After installation, a Salesforce administrator sets WatchDog up from the WatchDog tabs in Salesforce. This page covers the first settings to make and what each module shows.

First steps

  1. Assign the permission sets: WatchDog User to everyone who uses WatchDog, and WatchDog Setup Admin and WatchDog Monitor Hub Admin to the people who configure WatchDog and run monitoring. Setup and Monitor Hub open only with those admin permission sets, System Administrators included.
  2. Assign a package to the org in Setup → Plan. Until an org has a package, its feature tabs show "not in your plan". Setup always opens.
  3. Set the integration user in Setup → General. It needs a full Salesforce licence: an Integration-licence user can't read Apex debug logs.
  4. Set the polling frequencies in Setup → Jobs & Polling, then choose Save Changes. A module whose frequency isn't saved doesn't run.
  5. Choose who is monitored in Monitor Hub. Debug logs are collected only for users with an active Monitor Hub session.
  6. Connect your alert channels in Setup → Integrations: Slack, Jira or Azure DevOps.
  7. Optional: add your AI provider and key in Setup → AI Assistant.
Logs from users without a Monitor Hub session are deleted from Salesforce without being stored, by default, to keep your org's debug-log storage free. You can change this in Setup → Jobs & Polling.

Operations modules

Monitor Hub

Starts and stops monitoring sessions for chosen users, at a debug level and for a length of time you pick. WatchDog creates and renews the Salesforce trace flags, so collection doesn't stop when a flag would expire. It also shows the last and next poll for the org.

Hypercare

Switch Hypercare on in Monitor Hub for a go-live or a risky release. While it's on, each distinct Apex or Flow error opens one ticket in Jira or Azure DevOps, with who, when, where and what, plus an AI explanation if AI is set up. When the same error happens again, WatchDog adds a comment to that ticket instead of opening a new one. A Hypercare indicator shows on every WatchDog tab while it's active.

Log Explorer

Search the debug logs WatchDog has collected, including logs Salesforce has since deleted. Opening a log shows it as an execution tree with each method's duration, governor limits, SOQL and DML counts, callouts and exceptions. From a log you can create a ticket, download the raw log, or see matching Nebula Logger entries.

Pulse

Dashboards across all collected logs: errors and warnings grouped into issues (with a count, first and last seen, and the users affected), transaction health (median and 95th-percentile duration by entry point, with Sync Now to bring it up to date), limits and callouts, async Apex jobs, and log volume. A failed async Apex job also appears as an issue, since such jobs rarely leave a debug log.

Nebula Logger

If your org uses Nebula Logger, WatchDog detects it and reads its error and warning entries. They appear in Pulse alongside WatchDog's own issues and in Log Explorer for the matching request. The integration user needs Nebula's LoggerLogViewer permission set. Set how often it's read in Setup → Jobs & Polling.

Process Lens

Record a named business process, such as quote to order, while someone runs it. Process Lens captures the logs for that session, totals the work done, and compares the run with earlier runs of the same scenario. A CI/CD pipeline can start, stop and fetch a run's report using a per-org API key that your AWS administrator creates.

Know Your Org

An org health score with metadata findings, and a Users view that shows which users and channels (browser, API, background jobs) drive API calls, queries, CPU time and errors, by country and device.

Security modules

Login Intelligence

Every sign-in with user, IP address, country, device and application, on a map and a timeline. Alerts for sign-ins from countries you haven't allowed, spikes in failed sign-ins, and off-hours access in the user's own time zone. Countries, the map and country alerts need Look up sign-in locations with ipinfo.io turned on in Setup → Alerts; it's off by default because it sends each new sign-in IP address to ipinfo.io.

Audit Sentinel

Setup Audit Trail changes, rated by risk, on their own timeline, so high-risk administrative changes stand out.

Data Exfiltration

Report exports and data downloads by user, with risk ratings and trends. The Watchlist flags users for closer monitoring over a date range, for example during a notice period: any export by a watched user raises a critical alert. The geo-fence (exports from outside your allowed countries) needs sign-in locations turned on, as for Login Intelligence. Needs Salesforce Event Monitoring.

Security Posture

Configuration risk in the org, tracked over time so you can see whether it's getting better or worse.

Email Health

Bounce spikes, volume changes and deliverability problems in the org's outbound email.

Alerts, tickets and AI

Alerts go to the channels set in Setup → Integrations: Slack, Jira and Azure DevOps. Thresholds are set per org in Setup. AI analysis uses your own provider and key (OpenAI, Anthropic, Azure OpenAI, Amazon Bedrock or Google Gemini). It runs only when a ticket is created or a user asks for it, so you aren't charged for analysis you didn't request.

Retention and your data

Everything WatchDog collects is stored in your AWS account. Retention periods for logs, notifications and security data are set in Setup. See what leaves your account for the outbound connections WatchDog makes.