Gitguardian outage history

Status: Resolved Resolution has been validated. Affected components Secrets Detection API (Operational)

Status: Resolved System is back to normal operations. Affected components Webhooks (Operational) Secret Detection - Scan Check Run (Operational)

Status: Resolved Resolved. Following database query optimizations and reduced load from noisy background jobs, no further performance degradation has been observed. Metrics have been stable since deployment of the fixes. Affected components Secret Detection - Scan Check Run (Operational)

Status: Resolved This incident is now resolved. Following fixes deployed by our team, we have not observed further performance degradation, and metrics have been stable since. Affected components Secret Detection - Scan Check Run (Operational)

We are currently experiencing degraded performance in our public ingestion pipeline. This is caused by an ongoing outage of the GitHub API, which GitHub has acknowledged on their own status page. Our team is actively monitoring the situation. Because the root cause is on GitHub's side, resolution depends on GitHub restoring their API. Our pipeline is designed to resume automatically once GitHub recovers, and we will provide further updates as the situation develops.

GitHub has begun restoring their API, and our public ingestion pipeline is recovering. Real-time event ingestion and commit retrieval are resuming. We are continuing to monitor the situation to confirm full recovery and ensure the pipeline is processing normally. We will provide a final update once the incident is fully resolved.

GitHub has restored their API, and our public ingestion pipeline has fully recovered. Real-time event ingestion and commit retrieval are operating normally again. This incident was caused by an upstream GitHub API outage and is now resolved. Thank you for your patience.

We are currently investigating an issue where some customers may experience intermittent spikes of slowness when processing check runs.Our team deployed a fix and is actively monitoring the situation. We will provide updates as the situation progresses.We apologise for the inconvenience.

The fix has successfully been deployed and the situation should be back to normal. Sorry for the inconvenience.

Status: Resolved What happened:From June 1 12:00am UTC to June 1 08:45am UTC, one of the component serving the public monitoring experienced some delays and a growing backlog of unprocessed messages. Impact:No data was lost. The only impact for customers could have been a delay between the detection of a secret and the moment the corresponding incident was created in the customer's dashboard. The problem is now fixed, the faulty component was able to re-start and to process the backlog of messages.

Status: Resolved We deployed a fix today that addressed the check run issues. A brief database load increase was observed post-deployment with no impact on check run results.This incident is now resolved. Affected components Secret Detection - Scan Check Run (Operational)

Due to an availability issue of github.com Actions APIs (https://www.githubstatus.com/incidents/z3c6q056q332), some check runs have timed out on our side, resulting in a neutral status.

We are currently experiencing an incident on GPUB that started around yesterday, 10PM Monday 13th of March. An unexpected activity led to an overload of the queuing system of GPUB. We are deploying a fix to resume processing and catch up with the lag. In terms of data, some events will need to be reprocessed afterward, and we need to assess if some events cannot be processed again.

The fix has been deployed, the processing has resumed at full speed and the awaiting queue will be processing within the next hour.

This incident has been resolved.

Following a upgrade operation of a network component, GIM endpoints have been unavailable most of the time during this incident between 09:48 and 10:24 UTC. Fix has been deployed, situation is still under monitoring.

During deployment of GIM new version, The app was malfunctioning for a few minutes. The events that failed to evaluate during the incidents have been processed again.

We have identified a Network issue on Internal Repositories Monitoring that impacted Hook, Dashboard and API.

Following a DNS misconfiguration, GitGuardian Internal Monitoring platform was unavailable for about 20 minutes. Impact: No realtime scan (webhook): - GitHub events from recent GHE instances and Github.com were redelivered after the incident - Other events (BitBucket, GitLab) have been lost but we fully rescanned repositories when possible No API for secrets detection: CI and pre-commit hook failed. Dashboard was unavailable. Impacted entries: - api.gitguardian.com - dashboard.gitguardian.com - hook.gitguardian.com

Since October 17th, 2022, 12:28am (CEST) we are sending notifications by email for certain incidents already present in your workspace. We are currently investigating this incident.

This incident has been resolved.

On October 17th 2022, at around 5pm GMT, The queuing system in charge of handling Github Check Runs and Real Time Monitoring accumulated tasks that were slow to process and impacted the whole system. The size of the task queue started increasing for several hours straight, delaying processing for all customers. until it started decreasing again and went back to normal the day after at around 1am GMT, where processing time was back to normal. This event occurred again on October 18th 2022, at around 4pm GMT, but has been detected earlier and have been solved at 5pm GMT. Actions were taken to avoid this issue to happen again: It includes a better monitoring of those queues to improve time to detect an issue. The montoring of processing latencies on a end-user point of view will be improved too. More long terms actions are also planned, including implementing mechanisms to avoid accumulating latency in case of slow tasks accumulations.

We are continuing to work on a fix for this issue.

This incident has been resolved.

From Friday 9th Sept 6:30pm UTC to Monday 12th Sept 7:30 UTC, data collectors services were partially down due to an internal error. Github events have been correctly gathered into GitGuardian platform but processing of those events was down. There is no data loss but no alerts have been sent during this period of time. The processing has resumed correctly on Monday 12th Sept 7:30 UTC, the backdata from the week-end is being analyzed and should be caught up around 12pm. Update Monday 12th Sept. 10:25am UTC: Queues processing is slower than expected, ETA for catch up with the lag is postponed to at least to Tuesday morning (UTC), more precise ETA will be communicated when available. Update Monday 12th Sept. 2:05pm UTC: Slowness issues have been solved, queues processing has resumed at normal speed, ETA for catch up with the lag is Monday 12th Sept. 5:00pm UTC

This incident has been resolved.

A fix has been implemented and the systems are getting back to normal progressively.

Start of the issue.

We have an issue with our database. We are still evaluating the impact.

The incident has been resolved at 9h10 CEST.

The incident has been resolved.

We have detected a spike in errors on some endpoints of GitGuardian Internal Monitoring API. Our teams are investigating.