Large-Scale Infrastructure Maintenance for Service Recovery
Resolved
Oct 14 at 05:41pm WIB
We sincerely apologize for the significant disruption you have experienced with endpoint services, consoles, and related features since October 4, caused by hardware overload and performance throttling on our main inference node.
We would like to provide an update on the current status. Although intensive repair efforts have been made since the initial notification, the underlying root cause requires much more substantial infrastructure intervention and repairs than originally anticipated.
In light of this, we would like to inform you that our system will undergo an extensive period of large-scale maintenance.
This maintenance period will include upgrades and replacements of critical hardware components as well as architectural optimizations to ensure the long-term stability, performance, and reliability of our AI services, including API endpoints that support services such as Chatbot Arena.
Estimated Maintenance Duration:
Currently, we estimate that these comprehensive repairs will take a minimum of one (1) month, and possibly longer. We understand the significant impact this delay will have on your operations, and we are fully committed to restoring services as quickly as possible while prioritizing the quality of the restoration.
Service Impact:
During this period, all services that depend on the affected endpoints, including the administration console and key AI features, will experience total disruption or significant performance degradation.
We greatly appreciate your patience and understanding during this critical repair period. Our technical team is working at full capacity to prioritize the gradual restoration of services once the main infrastructure has been successfully upgraded.
Affected services