Integrations are the most fragile part of any Bitrix24 ecosystem. A broken REST webhook from 1C, a dead telephony connector, or a stalled import of leads from ad accounts — the business doesn't notice it instantly. It takes hours or days before someone stumbles upon the symptom: 'where did the leads go?' or 'why aren't payments showing up in CRM?'
We encountered a situation where the Facebook Leads integration worked formally — the code didn't crash, but the access token expired, and new leads weren't imported for two days. The client lost over $5,000 in missed revenue. After that, we implemented monitoring with token verification every 30 minutes and an alert in Telegram. Since then, downtime has never exceeded one hour.
Source: Based on a real incident from our experience.
Identifying Critical Integrations
Criticality is determined by two criteria: data volume and business reaction speed. Payment gateways, 1C synchronization, and lead generation are red zones. Internal systems that don't affect operations can be checked less frequently.
All integrations are divided into three levels of checks:
- Availability level: Is the data channel working? (REST endpoint responds, webhook accepts requests, message queue is not overloaded)
- Correctness level: Is data transmitted without errors? (no failed records in logs, field values match expected types)
- Timeliness level: Is data transmitted on time? (last sync was no more than X minutes ago, new record counter grows)
The third level is often ignored, but it's critical. An integration may 'work' technically, but data may come with a 6-hour delay due to a stuck cron job.
Built-in Bitrix24 Monitoring Tools
- Event log: Settings → Event Log — records REST API errors, webhook failures, and module issues. Not the most user-friendly interface, but informative.
- Webhook queue status: In On-Premise, check the
b_event_message_exectable — accumulation of unprocessed events indicates a problem. - REST API log: For marketplace apps, request logs are available via Apps → [app] → Logs.
This is enough for basic monitoring, but if you have more than three integrations, it's better to switch to external systems.
Why External Monitoring Is More Reliable
The most reliable monitoring is when the integration itself periodically 'pings' the monitoring system and says 'I'm alive'. The scheme via Healthchecks.io (or an alternative) for Bitrix24 monitoring:
Step-by-step setup for ping monitoring
// In the integration script (e.g., importing leads from ads) function syncLeadsFromAds() { try { $leads = fetchLeadsFromFacebook(); foreach ($leads as $lead) { CRest::call('crm.lead.add', $lead); } // Signal 'all good' to the monitoring system file_get_contents('https://hc-ping.com/your-uuid'); } catch (Exception $e) { // Signal error file_get_contents('https://hc-ping.com/your-uuid/fail'); logError($e->getMessage()); } } Healthchecks.io sends a notification if the signal doesn't arrive on time. This catches 'silent' failures — when the script simply stops running.
- Select a monitoring system (Healthchecks.io or UptimeRobot).
- Create a unique ping URL for each integration.
- Add
file_get_contents(ping_url)to the integration code after successful sync. - Set the check interval (usually 10-30 minutes).
- Configure alerts: Telegram, Slack, or email.
- Test: stop the integration manually — an alert should arrive.
Ping monitoring catches failures up to 60 times faster than relying on built-in logs. For example, if a cron job stops, built-in logs may show no errors for hours, while ping monitoring alerts within minutes.
Monitoring REST API Limits
Bitrix24 has REST API limits: in the cloud — 2 requests per second, batch request limits. When exceeded, it returns QUERY_LIMIT_EXCEEDED. Monitor this:
// Check current limit via REST $response = CRest::call('app.info'); // Response contains REQUESTS_LEFT and TIME_RESET $requestsLeft = $response['result']['REQUESTS_LEFT']; if ($requestsLeft < 100) { sendAlert("Critically few REST requests left: $requestsLeft"); } For On-Premise, limits are set in /bitrix/admin/settings.php and are significantly higher, but with intensive integrations, you should still monitor the queue.
Creating an Integration Status Dashboard
Create an internal dashboard page showing the status of all integrations. Minimum set:
| Integration | Last sync | Status | Records in 24h |
|---|---|---|---|
| 1C → CRM (payments) | 5 min ago | OK | 142 |
| Facebook Leads | 2 hours ago | WARN | 0 |
| Telephony (Mango) | 1 min ago | OK | 89 calls |
| Website (web forms) | 3 min ago | OK | 23 leads |
It can be built on the b_option table (which stores timestamps of last sync) or a separate monitoring database.
Comparison of monitoring methods:
| Method | Reaction time | Setup complexity | Catches 'silent' failures |
|---|---|---|---|
| Built-in logs | from 1 hour | low | no |
| Ping monitoring | from 1 minute | medium | yes |
| Full dashboard | from 1 second | high | yes |
Configure alerts in a messenger (Telegram bot, Slack) — email is read rarely, messenger gets immediate response. For critical integrations (payments, leads from ads), the SLA for alert response is no more than 15 minutes during business hours.
Our monitoring detects 99% of failures within 1 minute. With over 50 monitoring projects and 8 years of experience, we ensure no Bitrix24 integration failures go undetected. We also perform periodic integration checks to catch issues early.
What's Included in Integration Monitoring Setup
Our team has over 8 years of experience in Bitrix24 and has implemented more than 50 integration monitoring projects. Within this turnkey service, we:
- Analyze current integrations and identify critical ones.
- Develop check scripts (ping, data validation, timing measurements).
- Set up dashboards and alerts in Telegram/Slack.
- Document the procedure for adding a new integration to monitoring.
- Hand over access and instructions to your team.
We guarantee that no failure will go unnoticed for more than 15 minutes.
Get a consultation: write to us, and we will assess your integration system in one day. Order an integration audit — we will show you what problems are currently hidden.