PostHog Status

Resolved: Event and property definitions backfill has been completed, and new definitions are being generated in real time. System stability is being actively monitored.

Monitoring: The issue has been identified and resolved, new event and property definitions are now being created properly. We're re-processing historical data to ensure all definitions are available for use.

Investigating: The database failover was completed successfully, however it did not appear to be the root cause and new event names are still delayed. We are still investigating this

Identified: We have identified the root cause as a database issue, we will shortly be performing a database failover which may result in some ingestion delays. There will not be any data loss

Investigating: New event names (ie posthog.capture('new event name')) are not being created or are severely delayed in being created. That means that you can't select them in the interface

Resolved: All data ingestion has caught up and our systems are back to normal. Thank you for your patience.

Monitoring: We've mitigated the underlying issue in our datastore and are now catching back up. We will be totally caught up in ~8 hours. No events were lost in this incident and we have identified several steps we will take to prevent this from happening again. <3 Posthog

Investigating: We're still monitoring our application. Our lag is at 4 hours right now—i.e., events appear inside PostHog 4 hours after they're sent—while we continue recovering some of the affected infrastructure. We'll update this once we have more information.

No data has been lost as we're still collecting all the events.

Investigating: In an attempt to sort out the problem , we've restarted some of our Clickhouse nodes. For that reason, we're also seeing some delays in batch exports. All batch exports will complete once the nodes are back online.

Investigating: We're still seeing degraded performance when ingesting events and session replays. We're still working on tweaking our internal infrastructure to process the backlog.

There will be no data loss as we're still storing all of the events for later processing.

Monitoring: We're still monitoring our infrastructure. Ingestion delay has plateaued at around 45 minutes of delay between sending the event vs. you seeing the events inside our platform. It's important to notice that there's no data loss as everything is being captured appropriately.

Monitoring: We've identified the likely culprit behind the increased ingestion time. We're monitoring it closely and expect to see a decrease in our ingestion lag soon.

Investigating: We're experiencing some delay in ingesting new events. We're investigating.

Resolved: Event and property definitions backfill has been completed, and new definitions are being generated in real time. System stability is being actively monitored.

Monitoring: The issue has been identified and resolved, new event and property definitions are now being created properly. We're re-processing historical data to ensure all definitions are available for use.

Investigating: The database failover was completed successfully, however it did not appear to be the root cause and new event names are still delayed. We are still investigating this

Identified: We have identified the root cause as a database issue, we will shortly be performing a database failover which may result in some ingestion delays. There will not be any data loss

Investigating: New event names (ie posthog.capture('new event name')) are not being created or are severely delayed in being created. That means that you can't select them in the interface

Resolved: This incident has been resolved.

Monitoring: We’re seeing some ongoing impact from the "Data Processing Delays in US region" incident reported on January, 6th and a small number of users may see increased data duplication for January as a result.

We’re already in the process of cleaning this up and expect it to be resolved within the next 12 hours.

Resolved: The issue has been resolved.

Monitoring: We've fixed the issue and we are currently monitoring.

Investigating: We're experiencing an elevated level of API errors and are currently looking into the issue.

Resolved: All data ingestion has caught up and our systems are back to normal. Thank you for your patience.

Monitoring: We've mitigated the underlying issue in our datastore and are now catching back up. We will be totally caught up in ~8 hours. No events were lost in this incident and we have identified several steps we will take to prevent this from happening again. <3 Posthog

Investigating: We're still monitoring our application. Our lag is at 4 hours right now—i.e., events appear inside PostHog 4 hours after they're sent—while we continue recovering some of the affected infrastructure. We'll update this once we have more information.

No data has been lost as we're still collecting all the events.

Investigating: In an attempt to sort out the problem , we've restarted some of our Clickhouse nodes. For that reason, we're also seeing some delays in batch exports. All batch exports will complete once the nodes are back online.

Investigating: We're still seeing degraded performance when ingesting events and session replays. We're still working on tweaking our internal infrastructure to process the backlog.

There will be no data loss as we're still storing all of the events for later processing.

Monitoring: We're still monitoring our infrastructure. Ingestion delay has plateaued at around 45 minutes of delay between sending the event vs. you seeing the events inside our platform. It's important to notice that there's no data loss as everything is being captured appropriately.

Monitoring: We've identified the likely culprit behind the increased ingestion time. We're monitoring it closely and expect to see a decrease in our ingestion lag soon.

Investigating: We're experiencing some delay in ingesting new events. We're investigating.

Resolved: Event and property definitions backfill has been completed, and new definitions are being generated in real time. System stability is being actively monitored.

Monitoring: The issue has been identified and resolved, new event and property definitions are now being created properly. We're re-processing historical data to ensure all definitions are available for use.

Investigating: The database failover was completed successfully, however it did not appear to be the root cause and new event names are still delayed. We are still investigating this

Identified: We have identified the root cause as a database issue, we will shortly be performing a database failover which may result in some ingestion delays. There will not be any data loss

Investigating: New event names (ie posthog.capture('new event name')) are not being created or are severely delayed in being created. That means that you can't select them in the interface