As shown in the figure below, we found that the timeout rate in RTB breakout has reached 30%.
Is this timeout rate considered normal ?
Could this issue be related to the processing time on our DSP server, or is it potentially a network-related problem preventing AdX requests from reaching our server? (Currently, our all five endpoints are configured to point to a single server.)
![]() |
|
Feedback
How was our support today?
[2025-03-28 17:47:18Z GMT] This message is in relation to case "ref:!00D1U01174p.!5004Q02vHDRT:ref" (ADR-00297186)