Major incidents SaaS vendor changes, page 9

1Password reported a major Device Trust disruption affecting the USA/Global Device Trust Authentication Service, with possible device posture check delays or authentication failures. It started at 2026-09-09 20:43 UTC and was resolved at 2026-09-09 21:11 UTC, lasting 27 min. The vendor identified and rolled back the triggering change.

Vanta reported a major incident with GCP Firestore sync operations: syncs for GCP Firebase projects and databases failed, leaving some resource data stale in the Vanta app. It started at 2026-09-09 19:56 UTC and was resolved at 2026-09-10 21:36 UTC, lasting 25 h 39 min; Vanta worked with the GCP team but stated no cause or specific fix.

Salesforce reported a major Mule Soft service disruption affecting 1 instance of Cloud Hub20 Spaces-US-EAST-2. It started at 2026-09-09 19:13 UTC and was resolved at 2026-09-09 22:00 UTC, lasting 2 h 47 min; no cause or fix was stated.

Buildkite’s major disruption affected Web UI, REST API, Agent API and job queues, causing failed requests, delayed scheduling and artifact-upload issues. Started 2026-09-09 19:07 UTC; resolved 2026-09-10 01:54 UTC (6 h 46 min), with full customer recovery reported since 20:28 UTC. An autoscaling feedback loop overwhelmed the Redis cluster with connections.

Zapier reported a major outage affecting Zap History from 2026-09-09 17:51 UTC to 2026-09-09 19:41 UTC, lasting 1 h 49 min. Recent Zap runs could not be accessed or reviewed, but executions continued normally. Zapier implemented a fix and confirmed runs should be available in Zap History again.

Census reported a critical outage affecting the Census Sync Management UI from 2026-09-09 16:41 UTC to 2026-09-09 17:33 UTC (52 min). A backend change prevented Activation workspaces from loading; rolling it back restored access. Only UI components were impacted; sync runs were unaffected.

Make reported a major degradation in EU1 (eu1.make.com), causing possible instability in scenario executions, webhook processing, and platform access. It started at 2026-09-09 16:11 UTC and was resolved at 2026-09-10 10:29 UTC, lasting 18 h 17 min; operational metrics had returned to healthy levels by 09-09 16:59 UTC.

Some Acuity Scheduling customers saw a "SORRY" error when accessing their Client Scheduling Page; Squarespace rated the incident major. It was marked resolved, with a recorded start of 2026-09-09 15:26 UTC and resolution of 2026-09-08 15:28 UTC; the timestamps conflict, and duration is unknown.

Census reported a major Sync Management UI degradation from 2026-09-08 20:04 UTC to 2026-09-08 22:06 UTC, lasting 2 h 1 min. A recent change blocked some UI jobs, delaying notifications and preventing dataset previews, sync configuration fields, and connection tests from loading for some organizations. A fix was implemented, and the incident was resolved.

Confluent Cloud reported a major connectivity disruption affecting Kafka clusters in AWS us-east-1, starting 2026-09-08 20:01 UTC and resolved 2026-09-09 13:01 UTC (reported duration: 16 h 59 min). A fix was implemented at 09-08 20:30 UTC; connectivity was reported fully operational at resolution.

Miro reported a major outage of username-and-password sign-in starting 2026-09-08 19:21 UTC, lasting 19 h 2 min and resolved at 2026-09-09 14:24 UTC by removing password login. Miro directed customers to passwordless sign-in; SSO and social login were unaffected.

dbt Platform users were logged out of active sessions and temporarily unable to log in during a major incident starting 2026-09-08 18:30 UTC. The incident lasted 4 h 46 min and was resolved at 2026-09-08 23:16 UTC after a fix was deployed to prevent sessions from terminating.

Attio reported intermittent platform availability issues with major impact, starting at 2026-09-08 17:05 UTC and resolved at 2026-09-08 18:42 UTC; the reported duration was 1 h 36 min. The team deployed a fix and monitored the system, but did not state a root cause.

Astronomer reported a major Astro degradation: tasks remained queued on some worker queues because autoscaling resource names exceeding Kubernetes’ 63-character limit were rejected, preventing worker scaling for a very small number of customers. It started at 2026-09-08 12:33 UTC and resolved at 2026-09-08 20:09 UTC, lasting 7 h 35 min; a fix was implemented.

GitLab Duo had a major outage: starting new chats with AI catalog agents failed as websocket connections immediately dropped. The incident started at 2026-09-08 08:54 UTC and was resolved at 2026-09-08 09:57 UTC, lasting 1 h 3 min; GitLab reverted the merge request that introduced the failure, restoring custom/catalog agents.

Aircall reported a major incident causing no audio on inbound or outbound calls for some Singapore customers, starting 2026-09-08 05:50 UTC and resolved at 2026-09-08 08:51 UTC; the computed duration was 3 h 0 min. Aircall attributed the issue to a partner and implemented a routing change before confirming resolution.

Otter.ai reported a major incident affecting Otter joining Microsoft Teams meetings for some customers, starting 2026-09-07 22:51 UTC and resolved 2026-09-08 16:55 UTC, lasting 18 h 4 min. Otter deployed a fix and confirmed successful joining was restored.

Cursor reported a major Grok Bot outage from 2026-09-07 20:32 UTC to 2026-09-07 21:37 UTC, lasting 1 h 5 min. Some users could not reach their Grok Bot computers, and Recover or Reset could fail; data was not affected. A fix was implemented, and the incident was resolved.

Ably reported a major outage affecting access to documentation at ably.com/docs from 2026-09-07 11:25 UTC to 2026-09-07 11:32 UTC, lasting 7 min. Realtime services were unaffected; a fix was implemented at 11:28 UTC, and the incident was resolved.

Xero reported a major AU Xero Tax outage from 2026-09-07 01:49 UTC to 2026-09-07 04:16 UTC, lasting 2 h 26 min. An ATO services outage prevented customers from filing tax returns or STP filings, though tax return creation remained available. ATO implemented a fix, and Xero confirmed resolution.

PlanetScale reported a major connectivity outage affecting Vitess databases in GCP from 2026-09-05 18:39 UTC to 2026-09-05 18:53 UTC, lasting 14 min. The vendor identified a networking issue and deployed mitigation, restoring connectivity; it also stated it was making improvements to address root causes and detection.

Lovable reported a major partial outage affecting the Editor for some users, starting at 2026-09-05 18:31 UTC and lasting 27 min. It was resolved at 2026-09-05 18:58 UTC; Lovable stated that the upstream issue was resolved.

Salesforce reported a critical Platform service disruption affecting 1 instance, starting at 2026-09-05 10:30 UTC and resolved at 2026-09-05 12:00 UTC, lasting 1 h 30 min. Salesforce confirmed its mitigation succeeded and service was stable, with customers no longer impacted; the technical trigger and underlying cause remained under investigation.

Monday.com reported critical connectivity issues affecting AUS - Platform, EU - Platform, Platform, and US - Platform, starting 2026-09-05 08:31 UTC. A fix was implemented at 09-05 09:45 UTC; the incident was resolved at 2026-09-05 10:39 UTC, with a reported duration of 2 h 7 min.

Socket’s public API experienced elevated latency and error rates in a major-impact incident from 2026-09-05 07:21 UTC to 2026-09-05 09:04 UTC, lasting 1 h 43 min. Socket reported implementing a fix at 07:41 UTC and subsequently resolved the incident; no cause was stated.

Perplexity reported a major incident affecting some connectors from 2026-09-05 04:50 UTC to 2026-09-05 07:09 UTC, lasting 2 h 19 min. The issue was mitigated at 06:52 UTC, with impacted connectors monitored for recovery before resolution; no specific cause or fix was stated.

Salesforce reported a critical Service Disruption affecting Marketing Cloud Core Service on 1 instance. It started at 2026-09-05 01:21 UTC and was resolved at 2026-09-05 01:33 UTC, lasting 12 min.

Semgrep reported a major outage starting 2026-09-04 23:13 UTC: an incomplete 1.176.1 patch release prevented PyPI installations from using Pro Engine. It recommended installing 1.175.1 or 1.176.0 temporarily. Resolved at 2026-09-04 23:38 UTC after 24 min, Pro Engine could again be installed and used normally with Semgrep 1.176.1 via PyPI.

GitHub reported a major Copilot code review outage: affected reviews failed to complete or post comments. Tracked from 2026-09-04 20:39 UTC to 2026-09-04 22:26 UTC (1 h 47 min), the final report places failures from 20:04 UTC. An authentication-permissions change blocked review submissions to the GitHub API; reverting it restored service.

GitHub’s major Copilot code review incident ran from 2026-09-04 20:39 UTC to 2026-09-04 22:26 UTC (1 h 47 min). Affected pull request reviews failed to complete or post comments because an authentication-permissions change blocked submissions to the GitHub API; reverting the change restored normal operation.

Inngest reported a major function-execution degradation causing failed runs or delayed scheduling from incoming events, starting 2026-09-04 18:26 UTC and resolved at 2026-09-04 19:10 UTC (43 min). The vendor said it identified the cause without disclosing details; customers can replay failed runs in the UI.

Sourcegraph Cloud experienced increased latency and error rates in US Central, with major impact, starting 2026-09-04 17:33 UTC. The incident lasted 5 days 4 h and was resolved at 2026-09-09 21:36 UTC; Sourcegraph linked it to an ongoing incident with its underlying Google Cloud infrastructure.

Hosted Grafana in GCP US Central experienced a major outage: some stacks were unavailable or failed to load because reduced compute capacity in a cloud provider availability zone prevented startup. Other regions were unaffected. The incident ran from 2026-09-04 17:03 UTC to 2026-09-04 20:42 UTC (3 h 39 min), when the region was fully operational.

Expo reported a major degradation with elevated queue times for Android builds and other Linux workflows from 2026-09-04 16:05 UTC to 2026-09-05 07:10 UTC, lasting 15 h 4 min. GCP capacity constraints and VM provisioning issues caused delays; queued jobs continued progressing, and queues drained after GCP resolved the upstream issue.

dbt Labs reported a critical outage in US Cell 1 GCP, with increased API errors and loss of access to dbt; its final update said all services were down. The incident started at 2026-09-04 14:20 UTC and was resolved at 2026-09-04 15:51 UTC (1 h 31 min). The vendor said it implemented measures to prevent recurrence.

Mailgun reported a critical EU messaging outage affecting API, SMTP, inbound email processing, outbound delivery, events and logs, and stats and analytics. It started at 2026-09-04 14:16 UTC and was resolved at 2026-09-04 14:33 UTC, lasting 17 min; a fix was implemented by 09-04 14:21 UTC.

Some Descript transcription requests failed in a major incident starting 2026-09-04 13:53 UTC. Descript reported functionality fully restored at 2026-09-04 14:09 UTC and marked the incident resolved at 2026-09-04 17:00 UTC; the recorded incident duration was 3 h 7 min.

Snowflake reported a critical outage in AWS - Europe (Frankfurt): customers may have been unable to access or use services and features through Snowsight. It started at 2026-09-04 07:34 UTC and resolved at 2026-09-04 10:10 UTC, lasting 2 h 36 min. Snowflake restored service by rolling back an underlying infrastructure configuration update; a postmortem RCA is available on Snowflake Community.

Astronomer reported a major outage affecting authentication, deployment access and DAG visibility, with 401 and RBAC errors for multiple customers. It started at 2026-09-04 03:08 UTC and was resolved at 2026-09-04 06:07 UTC, lasting 2 h 58 min. A network configuration change during an infrastructure upgrade disrupted control-plane connectivity to customer deployments; Astronomer applied a fix and confirmed normal operation.

Descript’s critical avatar generation outage ran from 2026-09-04 01:31 UTC to 2026-09-04 05:17 UTC, lasting 3 h 46 min. Avatar generations failed due to a third-party outage; Descript reported successful generations after that service returned and marked the issue resolved.

Cursor reported major degradation affecting Grok Bot routine triggers: some trigger fires could be dropped. The incident started at 2026-09-03 23:42 UTC and was resolved at 2026-09-04 00:42 UTC, lasting 59 min. The vendor said it identified the root cause and was implementing a fix, but provided no specifics.

Semgrep reported a major degradation of SMS Managed Scans starting 2026-09-03 21:35 UTC, lasting 1 h 10 min and resolving at 2026-09-03 22:45 UTC. The vendor attributed the disruption to an AWS STS outage in us-west-2 and confirmed full recovery of all Semgrep products at resolution.

Perplexity reported a critical usage-based billing incident: Computer was degraded and incorrectly reported insufficient credits. It started at 2026-09-03 20:30 UTC and was resolved at 2026-09-03 21:10 UTC, with a reported duration of 39 min; a fix was implemented by 2026-09-03 20:44 UTC.

Segment reported a major outage affecting Engage UI page loads, including Audiences and Journeys, in the US. The incident started at 2026-09-03 18:30 UTC and lasted 4 h 11 min, resolving at 2026-09-03 22:41 UTC; a fix was applied at 09-03 20:26 UTC, restoring full UI functionality.

MongoDB Atlas reported major impairment of cluster operations in AWS me-central-1 and AWS me-south-1, affecting MongoDB Cloud, starting 2026-09-03 16:44 UTC. The incident remained open under monitoring when archived; the recorded open duration was 2 minutes, with no cause or fix stated.

JIRA’s critical US East disruption started at 2026-09-03 15:57 UTC and resolved at 2026-09-03 17:02 UTC (computed duration: 1 h 4 min), leaving affected users unable to access services. Recovery began at 16:15 UTC after mitigation; the postmortem attributes the disruption affecting Jira and Jira Service Management to a progressive rollout of a caching change.

Some Discord servers appeared unavailable in the client during a major outage starting 2026-09-03 14:35 UTC and lasting 1 h 2 min. Discord implemented a fix at 15:01 UTC and resolved the incident at 2026-09-03 15:37 UTC; no cause was stated.

Cursor reported a major incident affecting Automations, CLI, Cloud Agents, and IDE from 2026-09-03 14:17 UTC to 2026-09-03 16:31 UTC (2 h 14 min). An upstream Anthropic issue caused elevated errors or failed Agent turns for Fable 5.1, Fable 5, Opus 5, Opus 4.8, and Opus 4.6 requests; service was restored.

Salesforce reported a major B2 B2 CCommerce disruption affecting Commerce cloud sandbox storefronts across 84 instances, starting 2026-09-03 13:50 UTC and resolved at 2026-09-03 13:51 UTC (1 min). It later confirmed the listed instances and services remained healthy with no impact, calling the notification a false positive.

Salesforce reported a critical B2 CCore service disruption affecting Commerce cloud ODS sandbox storefronts on 1 instance. It started at 2026-09-03 13:50 UTC and was resolved at 2026-09-03 14:30 UTC, lasting 40 min; Salesforce restored access to the affected storefronts.