Sign in

Admin

Salesforce API Usage, Automation and Writes

These pages help an org admin watch the Salesforce connection: how much of your Salesforce API quota is in use, how many calls Forge makes each day, and whether writes Forge queued for Salesforce are keeping up. All of them need the org admin role.

Salesforce API usage (Admin > Salesforce API usage)

Salesforce gives every org daily and hourly allowances (limits) for API calls, bulk jobs, emails and similar. This page reads those limits live from Salesforce and refreshes itself every 60 seconds. Use it to find out whether Forge, or anything else connected to the org, is close to running out.

Key limits cards

Up to six cards at the top show the Salesforce limits Forge watches most closely: current used amount out of the maximum, a usage bar and how much remains. Forge tracks these limits: Daily API Requests, Daily Bulk API Requests, Daily Bulk V2 Query Jobs, Daily Bulk V2 Query File Storage MB, Daily Async Apex Executions, Daily Generic Streaming API Events, Daily Streaming API Events, Daily Workflow Emails, Hourly Async Report Runs, Hourly Dashboard Refreshes, Hourly Sync Report Runs, Hourly Time Based Workflow, Single Email, Mass Email, Streaming API Concurrent Clients, Concurrent Async Get Report Instances and Concurrent Sync Report Runs. Only limits your org reports are shown.

Warning banner and color coding

UsageColorWhat to do
Under 70%GreenNothing. Normal.
70% to 89%Amber. A banner at the top lists every tracked limit at or above 70%.Find out what is consuming the quota (see the daily Forge call table below and your other integrations) before it runs out.
90% or moreRedAct now. When Daily API Requests reaches its max, Salesforce rejects further calls, so Forge reads and writes can fail until the 24-hour window rolls over.

All limits table

Column or controlWhat it doesRecommendation
LimitSalesforce's name for the allowance. Click the header to sort.Sort by Usage to see the most-used limits first.
Used / Max / RemainingHow much of the allowance is consumed, the total, and what is left. All columns sort.Watch Remaining on Daily API Requests; it is the one most likely to affect Forge.
UsagePercent used, with the same color coding as the banner, plus a bar.Treat anything above 70% as worth investigating.
Filter boxNarrows the table to limits whose name or key contains the text you type.Type "email" to see Single Email and Mass Email limits together.
Show moreThe table shows 20 rows at a time; Show more reveals the next 20. Changing the filter resets to 20.Filter first if you are looking for one limit.

The page shows the time the data was last fetched at the bottom.

Forge calls to Salesforce (daily table)

Below the limits, a Day (UTC) and Calls table counts how many Salesforce calls Forge itself made for your org each day, with 90 days of history. Days are UTC, so the day boundary may not match your local midnight.

To work out whether Forge is the heavy API user: compare the Calls for a day with the Used amount on Daily API Requests. If Forge's calls are a small fraction, the rest is coming from other integrations, reports or data tools connected to the same org.

Salesforce Automation (Blackthorn staff only)

Salesforce Automation (/admin/sf-apex) holds a single org-wide switch, Forge Apex, that turns Forge's Apex triggers, callouts and REST endpoints in your org on or off. These sync registration data, attendee updates and session changes between Salesforce and Forge in real time. The page lists the affected Apex: the Attendee and Session Attendee Planner Notifier triggers, the Registration Notifier class, the Registration Trigger Queue, and the Health Check REST endpoint.

This page is hidden from customer orgs at GA: only Blackthorn staff can open it. If you believe Forge Apex needs to be paused in your org, contact Blackthorn support instead of looking for the switch.

Disabling Forge Apex stops automatic syncing between Salesforce and Forge. Registration notifications and attendee updates are not processed until it is enabled again. The page asks for confirmation each time you change it.

If the page reports "Custom setting not deployed", the Planner UI Settings custom setting is not in the Salesforce org yet; the Forge metadata package must be deployed before the switch exists.

Salesforce writes (Admin > Salesforce writes)

When you save something in Forge, the change is written to Salesforce. After a Salesforce write has already succeeded, Forge keeps a local copy of the record (a projection) so lists load quickly, and queues a follow-up job in the write outbox to confirm the copy matches. The outbox does not perform your write itself. Drift detection compares those local copies with live Salesforce data to catch records that changed outside Forge. This page is the health dashboard for both.

Summary cards

CardWhat it showsHow to read it
Write outboxNumber of rows queued or retrying, how many are dead, and how many minutes the oldest pending row has waited.A healthy org shows 0 or a low number that clears quickly. A growing count, or an oldest-pending age above your warning threshold, means the follow-up jobs are not keeping up.
Drift stateHow many projections are tracked and how many have drifted.Drifted is not an outage. It means a Salesforce record's last-modified time no longer matches Forge's copy, usually because someone edited it directly in Salesforce or another integration did.
Drift repair (24h)Drift repair runs in the last 24 hours, with counts that succeeded, failed and are currently active.Failed repair runs deserve a look at the Reason column in the drift table.

SLO thresholds

The SLO thresholds decide when a pending outbox row is considered late. Click Edit, change the two numbers, then Save (or Cancel).

FieldWhat it doesRecommendation
Warning (min)Minutes a row may stay pending before it counts as a warning. Whole number from 1 to 1440. Default shown by the page is 15.Keep the default unless you routinely queue large batches that legitimately take longer.
Critical (min)Minutes pending before it counts as critical. Whole number from 1 to 1440 and must be at least the Warning value. Default shown by the page is 60.Set it to the longest delay your team can tolerate before someone should be paged or investigate.

If a banner says the background processor hasn't run, rows are past the critical threshold with zero attempts: the scheduled worker is not picking them up. Retry on a row and Run drift check both process immediately, so you can use them to unblock things while the cause is investigated.

Write outbox table

Control or columnWhat it doesRecommendation
Status filterAll, Queued, Retrying, Processing, Done or Dead.Choose Dead to see rows that gave up; choose Retrying to see rows still trying.
Drain nowProcesses up to 100 pending rows right away instead of waiting for the background worker, then reports how many succeeded, are retrying, died and remain.Use it after a Salesforce outage or after raising an API limit, when you want the backlog cleared now.
Object / Record / OpThe kind of Salesforce record, its ID, and the operation (create or update).Use the Record value to find the record in Salesforce.
StatusQueued (waiting), Processing (running now), Retrying (failed and will be tried again later), Done, or Dead (stopped; no more automatic attempts).Dead needs a human decision; the other states resolve themselves.
Attempts, AgeHow many times it has been tried and how long ago it was created.Many attempts on a young row means a persistent error; read the Error column.
ErrorThe last error Salesforce or Forge returned.Typical causes are a deleted record, a validation rule, a permission change or an API limit.
RetryRuns that row again immediately. If the Salesforce lookup fails again you are told and the error is updated.Use after fixing the cause of the error.
KillMarks the row dead so it stops retrying.Use only when the record no longer needs syncing, for example it was deleted in Salesforce on purpose. Done rows have no actions.

Drift detection table

Control or columnWhat it doesRecommendation
State filterAll, Unknown, In sync or Drifted.Start with Drifted.
Run drift checkQueues a check of up to 100 projections against live Salesforce data. Refresh after a moment to see results.Run it after a bulk edit or data load done directly in Salesforce. The list is empty until a check has run.
StateIn sync (the Salesforce last-modified time matches), Drifted (it does not, or the record is missing), Unknown (not yet checked or repair queued).Drifted records are repaired by queuing a refresh write; the state returns to Unknown until the next check confirms.
Last op, CheckedThe last Forge operation on the record and when it was last verified.Old Checked dates mean the record has not been verified recently.
ReasonWhy a record is Drifted: missing_in_salesforce (record not found), last_modified_mismatch (changed in Salesforce since Forge last saw it), repair_queued (a repair write is waiting), or an error message from the check.missing_in_salesforce usually means a record was deleted outside Forge.

Tables show 20 rows with Show more, and the page refreshes every 30 seconds.

These pages are available to org admins only.