Pinpoint Outage History

Started Sep 01, 2026, 08:20 AM UTC· resolved Sep 01, 2026, 03:01 PM UTC

1. investigating Sep 01, 2026, 08:20 AM UTC

   We are currently investigating an issue affecting emails that are becoming stuck in a “Sending” status. Our engineering team is aware of the issue and is actively investigating. We’ll provide further updates as we learn more and work towards a resolution.

2. identified Sep 01, 2026, 09:40 AM UTC

   We have identified the issue - our engineering team is currently working on a fix and will provide a further update once this has been resolved.

3. monitoring Sep 01, 2026, 10:30 AM UTC

   A fix has been implemented and we are monitoring the results.

4. resolved Sep 01, 2026, 03:01 PM UTC

   This incident has been resolved.

5. postmortem Sep 02, 2026, 11:07 AM UTC

   ## **Email delivery disruption - 29 August to 1 September 2026** **What happened** Between Saturday 29 August and the afternoon of Monday 31 August, and again from 23:30 BST on Monday 31 August until 11:09 BST on Tuesday 1 September, emails sent from Pinpoint were not delivered. This affected candidate communications, notifications and other outbound email across the platform. **Cause** Both disruptions were caused by our email delivery provider, Postmark. An account restriction was applied on their side and, following its resolution, a required manual review by their team was not completed. This left the restriction in place and caused their systems to accept our emails as sent while not delivering them. No errors were returned to us, which is why the second disruption was not detected until customers reported it. Postmark has confirmed the fault was theirs, apologised, and made changes to their review process and our account settings to prevent a recurrence. No changes to Pinpoint contributed to the incident. **Resolution** On Tuesday morning our engineering team identified that the issue was upstream and restored delivery by routing all email through a new sending channel with Postmark. Delivery was confirmed working by 11:09 BST. All emails affected during the disruption have since been delivered. A small number of recipients may have received duplicate copies of some emails as a result of the recovery; we apologise for any confusion this caused. **What we're doing** * We have added the ability to switch email delivery channels immediately via configuration, so a similar provider-side issue can be worked around in minutes rather than hours. * We have expanded the notifications we receive from our provider so that account issues are visible to a wider group of our team, including out of hours. * We are adding monitoring on delivery volume so that a drop in delivered email is detected by us, independently of whether our provider reports an error. * We have formally raised the handling of this incident with Postmark and are reviewing our email delivery arrangements more broadly. We're sorry for the disruption this caused you and your candidates. If you have any questions or have identified any specific emails you believe were not delivered, please contact your Customer Success Manager or [[email protected]](mailto:[email protected]).

Read the full incident report →

Started Aug 03, 2026, 02:33 PM UTC· resolved Aug 03, 2026, 04:42 PM UTC

1. monitoring Aug 03, 2026, 02:33 PM UTC

   Our support provider, Fin (formerly Intercom), is currently experiencing a service outage. As a result, we're temporarily unable to respond to support messages. We're monitoring the situation closely via Fin's status page: https://www.finstatus.com/us-hosting, and we'll resume responding to your questions as soon as the service is restored. Thank you for your patience and understanding.

2. resolved Aug 03, 2026, 04:42 PM UTC

   Support Update – Resolved Fin (formerly Intercom) has deployed a fix, and the issue now appears to be resolved. Our team once again has access to the support inbox and is responding to your queries. We'll continue to monitor the situation closely in case anything changes. Thank you for your patience while this issue was being resolved.

Read the full incident report →

Started Jun 05, 2026, 12:33 PM UTC· resolved Jun 06, 2026, 12:30 AM UTC

1. monitoring Jun 05, 2026, 12:33 PM UTC

   We're currently experiencing an issue with our LinkedIn integration that may affect customers using RSC (Recruiter System Connect) features. You may notice delays in data syncing to Pinpoint, including in-mail messages and candidate exports. This will only impact customers using LinkedIn RSC features only. Core Pinpoint functionality is unaffected. We've identified the root cause and have a fix ready to deploy. We expect to roll this out this evening (UK time).

2. resolved Jun 06, 2026, 12:30 AM UTC

   This incident has been resolved.

Read the full incident report →

Started May 13, 2026, 06:29 AM UTC· resolved May 13, 2026, 07:04 AM UTC

1. investigating May 13, 2026, 06:29 AM UTC

   We are currently experiencing a service disruption caused by an issue with our primary database. While users are still able to log in and view data in Pinpoint, actions that involve saving or updating information are temporarily unavailable. Our team has identified the root cause and is actively working to restore full functionality. We will provide further updates as the situation progresses.

2. monitoring May 13, 2026, 06:54 AM UTC

   A fix has been implemented and we are monitoring the results.

3. monitoring May 13, 2026, 06:58 AM UTC

   We are continuing to monitor for any further issues.

4. resolved May 13, 2026, 07:04 AM UTC

   This incident has been resolved.

5. postmortem May 13, 2026, 02:16 PM UTC

   **Duration:** Approximately 4 hours of intermittent disruption \(03:00 – 06:51 UTC\), with two distinct impact windows. ## What Happened Pinpoint experienced an unexpected service disruption in the early hours of 13 May. The incident had two impact windows: an initial period of unavailability between approximately 03:00 and 04:00 UTC, and a follow-on period between approximately 05:39 and 06:51 UTC during which users attempting to modify certain types of records \(predominantly scorecard submissions\) encountered errors. No data was lost or corrupted, and no background jobs were delayed. ## Timeline \(UTC\) * 03:00 Initial impact begins; application unavailable. * 04:00 Initial impact resolves; engineering team investigating. * 05:39 Second impact window begins; errors on a subset of record updates. * 06:29 Status page warning posted. * 06:51 Second impact window resolves. * 07:04 Incident marked as resolved. ## Why It Happened The disruption was caused by an unexpected interaction between a routine internal database operation and an active background process. This combination of events is an outlier we had not previously encountered or anticipated. The root cause has been fully investigated and is understood. ## What We're Doing About It * **Additional Monitoring** We have put additional monitoring in place to detect this specific class of interaction much earlier, before it can cause customer impact. * **Increased Database Capacity** We have provisioned additional database capacity to further reduce the likelihood of recurrence. _We sincerely apologise for the disruption and appreciate your patience._

Read the full incident report →

Started Mar 19, 2026, 12:01 PM UTC· resolved Mar 19, 2026, 12:21 PM UTC

1. investigating Mar 19, 2026, 12:01 PM UTC

   We’re aware of a current issue causing 503 errors and are investigating urgently.

2. resolved Mar 19, 2026, 12:21 PM UTC

   This has been resolved. A full post mortem will follow soon.

3. postmortem Mar 19, 2026, 01:37 PM UTC

   **Date:** 19 March 2026 **Duration:** Approximately 35 minutes \(11:46 to 12:21 UTC\). ### What Happened Pinpoint experienced a service outage beginning at approximately 11:46 UTC on 19 March. Users were unable to access the platform and received 503 errors. Full service was restored by approximately 12:21 UTC. No data was lost or corrupted. ### Timeline \(UTC\) * **11:46** - Automated monitoring alerts triggered for service unavailability. * **11:47** - Engineering team began investigating. * **12:01** - Status page updated to reflect the incident. * **12:10** - Root cause identified as a database firewall misconfiguration introduced during an infrastructure change. * **12:20** - Fix applied; service restored. * **12:21** - Incident marked as resolved. ### Why It Happened During an infrastructure configuration change, a database firewall rule was inadvertently applied with an overly restrictive scope. This prevented our application servers from connecting to the primary database, causing all requests to fail. ### What We're Doing About It * **Immediate fix:** The firewall configuration has been corrected and verified. * **Hardening:** We are making production database firewall rules explicitly immutable and tightening API token permissions to prevent unexpected firewall modifications. We sincerely apologise for the disruption and appreciate your patience.

Read the full incident report →

Started Mar 18, 2026, 08:22 AM UTC· resolved Mar 18, 2026, 10:02 AM UTC

1. identified Mar 18, 2026, 08:22 AM UTC

   We are aware of an issue affecting the resume/CV preview within the platform. Resumes may appear blank or display a browser blocking message when viewed inline. **Downloading resumes is unaffected** — you can continue to access candidate resumes using the download button. Our team has identified the cause and a fix is being deployed soon. We expect normal service to be restored shortly. We apologise for the inconvenience and will update this page once the fix is live.

2. monitoring Mar 18, 2026, 09:16 AM UTC

   A fix has been implemented and we are monitoring the results.

3. resolved Mar 18, 2026, 10:02 AM UTC

   > What happened Users were unable to preview CVs and cover letters within the platform. The preview area appeared blank or showed a "blocked by browser" message. Downloading resumes continued to work normally throughout. > What caused it A security update was deployed overnight to protect admins from potentially harmful files. While the protection worked as intended, it had an unintended side effect - it also prevented the resume preview from loading correctly in all browsers. > What we did Identified the cause within minutes of being alerted Deployed a fix to restore normal resume previewing > What's next We have applied an alternative security protection using a different approach that doesn't affect the preview.

Read the full incident report →

Started Mar 03, 2026, 05:00 AM UTC· resolved Mar 03, 2026, 07:01 AM UTC

1. resolved Mar 03, 2026, 10:47 AM UTC

   Component: Pinpoint → Partial Outage Investigating - 05:00 UTC: > We are aware that some users may be unable to perform actions that involve saving or updating data. Our engineering team is actively investigating. Identified - 06:45 UTC: > The auto-scaling configured on our primary database infrastructure failed to provision additional storage in response to increased load. As a result, the database entered read-only mode, preventing write operations. We are provisioning additional storage to restore full functionality. Monitoring - 06:57 UTC: > Additional storage has been provisioned and write operations have been restored. Our team is monitoring to confirm full stability across all services. Resolved - 07:01 UTC: > The issue has been fully resolved and all services are operating normally. We apologise for the disruption and are reviewing our auto-scaling configuration to prevent recurrence.

Read the full incident report →

Started Feb 12, 2026, 02:05 PM UTC· resolved Feb 12, 2026, 03:02 PM UTC

1. investigating Feb 12, 2026, 02:05 PM UTC

   We are currently experiencing a full service outage. All requests are returning 503 errors. Our engineering team is actively investigating and working to restore service as quickly as possible.

2. investigating Feb 12, 2026, 02:09 PM UTC

   We are currently experiencing a major service disruption. Most requests are returning 503 errors, though limited service is available through a single server. Our engineering team is actively investigating and working to restore full capacity.

3. identified Feb 12, 2026, 02:18 PM UTC

   We are continuing to recover from a service disruption that caused 503 errors. Additional servers have been brought back online and service is improving. Our engineering team is working to restore full capacity.

4. monitoring Feb 12, 2026, 02:26 PM UTC

   Service has been mostly restored following an earlier outage. Some users may experience minor intermittent slowness. Our team continues to monitor as we return to full normal operation.

5. monitoring Feb 12, 2026, 02:46 PM UTC

   Service has stabilised following the earlier outage. Our servers are back online and performing as expected. Our team will continue to monitor the situation.

6. resolved Feb 12, 2026, 03:02 PM UTC

   The earlier service outage has been resolved. Our servers are back online and operating normally. We apologise for the disruption and are taking steps to prevent this from recurring.

7. postmortem Feb 13, 2026, 10:53 AM UTC

   **Date:** 12 February 2026 **Duration:** Approximately 30 minutes \(14:00 - 14:35 UTC\), with a brief secondary disruption during recovery. ### What Happened Pinpoint experienced a service outage beginning at approximately 14:00 UTC on 12 February. Most users were unable to access the platform and received 504 timeout errors, though partial capacity meant some requests were still being served. Full service was restored by approximately 14:35 UTC. Some outbound emails were delayed due to background workers also being affected. All queued emails were delivered by approximately 16:15 UTC. No data was lost or corrupted. ### Timeline \(UTC\) * **14:00** - Automated monitoring alerts triggered for service unavailability. * **14:02** - Engineering team began investigating. * **14:08** - Root cause identified as a network configuration conflict introduced by a third-party infrastructure provider. * **14:20** - Initial fix applied; service stabilising. * **14:22** - Third-party provider's automation reinstated the conflicting configuration, causing a brief secondary disruption. * **14:30** - Fix reapplied with additional safeguards to prevent recurrence. * **14:35** - Full service restored. * **16:15** - All delayed outbound emails delivered. ### Why It Happened A third-party infrastructure provider applied an automated network configuration change to our servers. The change was overly broad and conflicted with our own networking configuration, causing the majority of inbound traffic to be dropped. During initial recovery, the provider's automation re-applied the conflicting change before we implemented stronger protections. ### What We're Doing About It * **Immediate Safeguards:** Protections are in place to prevent this specific change from affecting our infrastructure again. * **Removing the Dependency:** We are accelerating work to bring the affected networking functionality fully in-house, removing the third-party dependency entirely. We sincerely apologise for the disruption and appreciate your patience.

Read the full incident report →

Started Nov 28, 2025, 12:03 PM UTC· resolved Nov 30, 2025, 07:30 PM UTC

1. investigating Nov 28, 2025, 12:03 PM UTC

   We’re currently investigating reports of missing records (including interviews, comments, messages and scorecards) as a result of the earlier reported infrastructure incident. The platform is operating normally, and no further service disruption is expected. Our team is actively working to identify and restore the affected data. We’ll provide updates as soon as we have more information. Thank you for your patience.

2. resolved Nov 30, 2025, 07:30 PM UTC

   We have identified the missing records within the platform and restored the affected data. If you have any questions or require assistance please reach out to our support team via the in-app chat function

Read the full incident report →

Started Jul 30, 2025, 11:22 AM UTC· resolved Jul 30, 2025, 02:36 PM UTC

1. monitoring Jul 30, 2025, 11:22 AM UTC

   Earlier today, our platform experienced a temporary disruption due to a sudden and highly unusual surge in external traffic, significantly beyond anything we’ve seen before. This caused intermittent 503 errors and degraded performance for approximately 15 minutes. We acted quickly to contain the issue and restore full service, including scaling our infrastructure and working closely with our edge network provider to bolster caching and traffic protection systems. Service is now fully operational and stable. We’re continuing to implement further safeguards to ensure that similar patterns of abnormal traffic cannot affect platform availability in the future. While today’s disruption was limited in scope and duration, we recognise the importance of uninterrupted access to Pinpoint - and are treating this with the seriousness it deserves. We’ll continue monitoring closely and share any further updates here as needed.

2. resolved Jul 30, 2025, 02:36 PM UTC

   This incident has been resolved.

Read the full incident report →

Started Jul 22, 2025, 06:38 AM UTC· resolved Jul 23, 2025, 07:07 AM UTC

1. investigating Jul 22, 2025, 06:38 AM UTC

   We experienced an unexpected outage with our primary database between 01:50 22nd July & 03:40 22nd July. Service has now been restored using a backup. However, the backup we restored from was approximately 24 hours old, which means any data written to the system between 01:00 BST 21st July and 01:00 22nd July is currently missing. We are in the process of restoring our database to a more recent back-up snapshot (from 01:00 BST 22nd July) so that the aforementioned missing data from yesterday is restored. Once this second restore has finished, any records created since 01:00 BST 22nd July will be missing from the system. We are actively working to ensure that this small amount of missing data from today is then also restored. All missing data is recoverable, it currently just exists in separate database back-ups. Once our second restore finishes the final part of our restoration involves reconciling the two back-ups. This process is complex and may take some time to complete. We will continue to provide further updates as progress continues. Service Availability: Fully restored. Data: Any records created or updated during the affected period (last 24 hours) are not yet visible. Next Steps: We are working to recover and restore this missing data. We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

2. investigating Jul 22, 2025, 08:07 AM UTC

   Database Outage Update: Recovery in Progress, Data Backfill Required We are still in the process of restoring our database to a more recent back-up snapshot (from 01:00 BST 22nd July) so that the aforementioned missing data from yesterday is fully restored. Once this second restore has finished, any records created since 01:00 BST 22nd July will still be missing from the system. We will then proceed with restoring this small amount of missing data from today. All missing data is recoverable, it currently just exists in separate database back-ups. Once our second restore finishes the final part of our restoration involves reconciling the two back-ups. Service Availability: Fully restored. Data: Any records created or updated during the affected period (last 24 hours) are not yet visible. Next Steps: We are working to recover and restore this missing data. We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

3. investigating Jul 22, 2025, 08:57 AM UTC

   Database Outage Update: Recovery in Progress, Data Backfill Required We have now restored our database to a more recent back-up snapshot (from this morning at 01:00 BST, 22nd July). Any and all records that were in any way created or modified yesterday (from 01:00 BST, 21st July onwards) are now visible once again in the system. We are now working to restore all data that was created or modified from 01:00 BST 22nd July (early hours of this morning) up until the point we restored our database just moments ago. All missing data is recoverable, it currently just exists in separate database back-ups, and we are working as fast as we can to reconcile the two back-ups. Service Availability: Fully restored. Data: Any records created or updated during the affected period (last 24 hours) are now visible. Next Steps: We are working to recover and restore data that was created or modified since 01:00 BST, July 22nd this morning. Important Note on Performance: - During this recovery process, system performance will be slower than usual. This is because we are temporarily running all operations through a single database while we work to fully restore normal infrastructure. - We are also experiencing significant slowdowns with the processing of all custom reports We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

4. investigating Jul 22, 2025, 10:12 AM UTC

   Database Outage Update: Recovery in Progress, Data Backfill Required We are now working to restore all data that was created or modified from 01:00 BST 22nd July (early hours of this morning) up until the point we restored our database at 08:57 BST today. All missing data is recoverable, it currently just exists in separate database back-ups, and we are working as fast as we can to reconcile the two back-ups. Service Availability: Fully restored. Data: Any records created or updated during the affected period (last 24 hours) are now visible. Next Steps: We are working to recover and restore data that was created or modified since 01:00 BST, July 22nd this morning. Important Note on Performance: - We have returned to normal database infrastructure, meaning the previously reported performance slowdowns should begin to improve load is shared across multiple databases. - We are also experiencing significant slowdowns with the processing of all custom reports We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

5. investigating Jul 22, 2025, 12:29 PM UTC

   Database Outage Update: Recovery in Progress, Data Backfill Required We are still working on restoring all data that was created or modified from 01:00 BST 22nd July (early hours of this morning) up until the point we restored our database at 08:57 BST today. All missing data is recoverable, it currently just exists in separate database back-ups, and we are working as fast as we can to reconcile the two back-ups. Service Availability: Fully restored. Data: Any records created or updated during the affected period (last 24 hours) are now visible. Next Steps: We are working to recover and restore data that was created or modified since 01:00 BST, July 22nd this morning. We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

6. investigating Jul 23, 2025, 06:20 AM UTC

   Database Outage Update: Recovery in Progress, Data Backfill Required Most of the data that was created or modified between 01:00 BST 22nd July and 09:00 BST 22nd July has been restored and is now visible in the platform. This includes all records with the exception of new job applications, which we are currently processing. We expect that data to be restored within the next hour. Service Availability: Fully restored. Data: Any records created or updated during the affected period (between 01:00 BST 22nd July and 09:00 BST 22nd July) with the exception of job application data is now visible in the system. Next Steps: We are now restoring remaining job application data that was created or modified between 01:00 BST, July 22nd and 09:00 BST. We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

7. resolved Jul 23, 2025, 07:07 AM UTC

   Database Outage Update: Recovery in Progress, Data Backfill Required All remaining data that was created or modified between 01:00 BST 22nd July and 09:00 BST 22nd July has been restored and is now visible in the platform. Service Availability: Fully restored. Data: Any records created or updated during the affected period (between 01:00 BST 22nd July and 09:00 BST 22nd July) is now visible in the system. We sincerely apologise for the inconvenience this has caused and appreciate your patience while we work to fully resolve the issue.

Read the full incident report →

Started Jul 22, 2025, 01:33 AM UTC· resolved Jul 23, 2025, 08:29 AM UTC

1. investigating Jul 22, 2025, 01:33 AM UTC

   We are investigating an outage with our primary database that is affecting the whole platform.

2. monitoring Jul 22, 2025, 02:40 AM UTC

   A fix has been implemented and we are monitoring the the application. You may experience degraded performance

3. resolved Jul 23, 2025, 08:29 AM UTC

   The outage has been resolved, and all services are operational. A full post-mortem will follow.

4. postmortem Jul 23, 2025, 08:29 AM UTC

   **Date:** 22 July 2025 **Duration:** 8 hours 34 minutes \(01:07 – 09:41 BST\) ### **What Happened** Pinpoint experienced a significant service disruption affecting our primary database, admin/API services, and careers site from 1:07 AM to 9:41 AM BST on 22 July. During this period, customers experienced complete service unavailability for approximately 2 hours and 45 minutes, followed by functionality with temporarily outdated data until full restoration was completed. ### **Timeline \(BST\)** * **01:07** – Service disruption began; database connectivity issues detected * **01:09** – Emergency response initiated; services marked as unavailable * **01:45** – Initial database restoration commenced * **03:45** – Services restored with backup data from the previous day * **09:21** – Second restoration initiated with more recent backup * **09:41** – Full service restoration completed with current data * **10:00\+** – Performance optimisation measures implemented * **23:50\+** – Restoration of some delayed data from 3:00-9:41 AM ### **Why It Happened** The incident occurred during a trial of a new database vendor's migration service. While testing their system, operations intended for a test environment were unexpectedly redirected to our production database, resulting in the unintended removal of data. The database vendor's system includes advanced features designed to help companies migrate data safely. However, the behaviour of these features during our testing phase was not clearly indicated in their interface. This led to test operations affecting our live production data instead of remaining isolated in the test environment, where they belonged. ### **What We're Doing About It** * **Immediate Recovery Completed:** All services have been fully restored. We are working to restore data created during the early morning hours of Tuesday. * **Enhanced Monitoring:** We are implementing additional safeguards to detect and prevent unintended database operations, including alerts for any bulk data modifications. * **Vendor Collaboration:** We are working closely with the database vendor to improve their user interface and documentation, ensuring that system behaviour is communicated during all operational phases. We apologise for this disruption to your service. We understand the critical nature of your recruitment processes and the impact this may have had on scheduled interviews and daily operations.

Read the full incident report →

Started Jun 26, 2025, 03:16 PM UTC· resolved Jun 26, 2025, 06:54 PM UTC

1. investigating Jun 26, 2025, 03:16 PM UTC

   We're currently investigating an issue affecting the Candidate Search functionality. Users are experiencing errors with no results returned when attempting to search for candidates. This only affects the All Candidates page. Our engineering team is actively working to identify the root cause and restore full service as quickly as possible. We’ll provide an update as soon as more information is available. We apologise for the inconvenience and appreciate your patience.

2. identified Jun 26, 2025, 03:21 PM UTC

   We've identified the root cause of the Candidate Search issue and have a fix prepared. The patch will be deployed within the next 60 minutes. We'll continue to monitor progress closely and provide an update once the fix is fully in place and search functionality is confirmed to be stable. Thank you for your continued patience.

3. monitoring Jun 26, 2025, 04:40 PM UTC

   We’ve deployed a fix for the candidate search issue and candidate search functionality has been restored. Our team is continuing to monitor the system closely.

4. resolved Jun 26, 2025, 06:54 PM UTC

   This incident has been resolved.

Read the full incident report →

Started May 01, 2025, 02:31 PM UTC· resolved May 01, 2025, 04:19 PM UTC

1. investigating May 01, 2025, 02:31 PM UTC

   We are currently investigating the issue.

2. investigating May 01, 2025, 02:45 PM UTC

   We have identified an issue with our primary database cluster and are working with our infrastructure provider to determine the root cause.

3. identified May 01, 2025, 03:04 PM UTC

   The issue has been identified and a fix is being implemented

4. monitoring May 01, 2025, 03:21 PM UTC

   A fix has been rolled out. All services are currently operational and we are continuing to monitor.

5. resolved May 01, 2025, 04:19 PM UTC

   This has been resolved. A full post mortem will follow soon.

6. postmortem May 02, 2025, 03:21 PM UTC

   **Date:** 1 May 2025 **Duration:** 50 minutes \(15:11 – 16:01\) ### **What Happened** Pinpoint experienced a complete admin/API service outage and a partial careers site outage, lasting from 3:11 PM to 4:01 PM BST on 1 May. During this 50-minute window, users could not access the platform and were presented with error messages when attempting to log in or use any part of the service. ### **Timeline \(BST\)** **15:11** – Monitoring alerts triggered for a service interruption. **15:13** – Root cause identified as database connection failures. **15:20** – Contacted managed database provider; issue confirmed as infrastructure-related. **15:45** – Provider identified a disk controller failure and began working on remediation; we started a cluster migration ourselves in case of further provider issues. **15:50** – We escalated the issue with the provider due to a lack of failover activation. **16:00** – Provider migrated the database to new hardware; connectivity restored. **16:01** – Full service restoration confirmed across all components. ### **Why It Happened** The root cause was a hardware failure on the part of our managed database provider. Specifically, a critical disk controller in their storage infrastructure failed, resulting in incorrect disk space reporting at the database level. This led our primary database instance to believe it had run out of storage space, and it began rejecting all connections, including those from our application servers. Although our agreement with the provider includes premium-grade redundancy and automated failover mechanisms, these safeguards failed to activate due to an undetected fault in their failover detection logic. As a result, traffic was not redirected to the standby replica, and manual intervention from their engineering team was required to restore normal operations. Unfortunately, this intervention took longer than acceptable, contributing to the total duration of the outage. ### **What We’re Doing About It** * **Hardware Migration Completed:** Our provider has successfully migrated our primary database to a new, healthy hardware environment. Service has been fully restored and is operating normally. * **Evaluating New Providers:** We chose our current provider for their strong reputation and reliability, but in light of this incident, we’re reassessing our options to ensure our infrastructure meets the highest standards of resilience and support. * **Working with Current Provider:** We’re working with our current provider’s engineering and account teams to understand why their failover system did not function as promised and to ensure such a failure does not recur prior to us migrating to an alternative provider. We sincerely apologise for the disruption and appreciate your patience as we work to strengthen our platform's stability.

Read the full incident report →

Started Mar 18, 2025, 10:05 AM UTC· resolved Mar 18, 2025, 12:18 PM UTC

1. investigating Mar 18, 2025, 10:05 AM UTC

   During routine maintenance to our search infrastructure, an update to the indexed data did not apply correctly, resulting in the inability to search for job applications in the admin application.

2. identified Mar 18, 2025, 10:05 AM UTC

   The issue has been identified and a fix is being implemented.

3. monitoring Mar 18, 2025, 10:58 AM UTC

   A fix has been implemented and we are monitoring the results.

4. resolved Mar 18, 2025, 12:18 PM UTC

   This incident has been resolved.

Read the full incident report →

Started Feb 16, 2025, 04:40 PM UTC· resolved Feb 18, 2025, 04:29 PM UTC

1. identified Feb 16, 2025, 04:40 PM UTC

   During routine maintenance to our search infrastructure, an update to the indexed data did not apply correctly, resulting in the inability to search for some resources in the admin application, such as candidates.

2. identified Feb 17, 2025, 10:41 AM UTC

   We have identified the cause of the issue and are steadily rolling out the fix.

3. identified Feb 18, 2025, 11:02 AM UTC

   We have restored functionality to 95% of searchable resources and continue to roll out the fix for the remainder.

4. resolved Feb 18, 2025, 04:29 PM UTC

   We have finished rolling out the fix for our search backend, and all search components are 100% functional again.

5. postmortem Feb 27, 2025, 03:32 PM UTC

   We apologise for last week’s disruption to our typically swift and flexible search functionality. This was caused by routine maintenance of our search infrastructure over the weekend of February 15th and 16th, which encountered unexpected issues during data migration. The resolution of the issue was further complicated by the larger-than-anticipated size of some datasets and the difficulty of certain technical processes in keeping up. On Tuesday, February 18th, we also deployed several bug fixes that sped up our data processing and ensured that search failures did not make the UI unusable in some circumstances. Rest assured, the team has thoroughly reviewed the incident, and changes are already in the pipeline to prevent similar occurrences.

Read the full incident report →