Resolved -
All replicated storage volumes are fully synchronised again between our provider's two datacenters, and normal operation has been confirmed. Platform performance is back to its usual level, and backup operations have resumed normally.
What happened: The incident originated from a loss of network redundancy between our hosting provider's two datacenters, following maintenance work on one of the inter-site links. The high-availability mechanisms worked as designed: affected virtual machines automatically moved to the second datacenter and restarted there, with storage remaining available throughout. No data was lost.
The trade-off during that period was performance: with redundancy reduced, storage and network response times degraded, which is what you experienced as slow navigation and occasional failures on heavy operations such as large queries or full searches. Restoring redundancy and resynchronising data between the two sites then took several days, during which occasional slowdowns continued.
Current status: All systems operational. We are continuing to monitor performance closely over the coming days.
Our provider has shared a detailed root cause analysis with us, and corrective measures are being put in place on their side to prevent a recurrence.
We sincerely apologise for the disruption over the past few days and thank you for your patience throughout.
Aug 21, 08:17 CEST
Update -
Storage response times returned to a normal level yesterday early afternoon and have remained stable since. The network link between our provider's two datacenters continues to operate normally.
What happens next: Data replication between the two sites is still being rebuilt, in successive batches under continuous monitoring so that platform performance is preserved. The process is progressing as planned, and a full return to normal operation is expected during the course of tomorrow. Until then, occasional short slowdowns may still be noticed.
What this means for you:
* The platform is available, and response times are at a normal level.
* Brief slowdowns may still occur until replication completes.
* If a heavy operation (large query, full search) fails, retrying a few minutes later usually works.
* We are keeping this component marked as degraded until the operation is fully complete.
If you notice persistent slowness or unusual behaviour in the application, please contact our support team so we can investigate and escalate it with priority.
Next update: tomorrow, or earlier if the situation changes significantly.
Aug 18, 11:46 CEST
Monitoring -
The network link between our provider's two datacenters was fully restored this morning, and traffic is now balanced normally across both connections.
During the incident, storage systems were responding more slowly than usual, with the most noticeable slowdown occurring this morning between approximately 10:00 and 11:50 CEST. Since 12:30 CEST, response times have returned to a normal level and remain stable.
What happens next: Data replication between the two sites is being rebuilt, starting this afternoon. This is done in successive batches under continuous monitoring in order to keep the impact on the platform as low as possible. The process is currently estimated to take 24 to 48 hours.
What this means for you:
* The platform is available, and response times have returned to a normal level.
* We are keeping this component marked as degraded until the resynchronisation is fully complete.
* If a heavy operation (large query, full search) fails, retrying a few minutes later usually works.
If you notice persistent slowness or unusual behaviour in the application, please contact our support team so we can investigate and escalate it to our provider with priority.
Next update: Tuesday 18 August, before 10:00 CEST — or earlier if the situation changes significantly.
Aug 17, 16:50 CEST
Update -
Our hosting provider has restored the second network connection between its two datacenters. The systems are now rebalancing data across both sites.
Work is still ongoing to bring performance back to its normal level. During this recovery phase, you may notice temporary slowdowns again — this is expected, and our provider is working to keep the impact as limited as possible.
In practice, this means:
* The platform remains available and usable.
* Navigation may still feel slower than usual at times.
*Heavy operations (large queries, full searches) may occasionally fail — retrying a few minutes later usually works.
We will keep this page updated as the situation evolves.
Aug 17, 13:42 CEST
Investigating -
Our hosting provider is currently experiencing redundancy issues between its two datacenters.
As a result, we are seeing performance degradation on our shared platform:
* Navigation within the application is slower than usual.
* Resource-intensive operations — such as heavy queries or full-text searches — may fail or time out.
Our provider is actively working on the issue. We will post an update as soon as we receive new information from them.
We apologise for the inconvenience and thank you for your patience.
Aug 17, 11:28 CEST
