SebastianGruza commented on PR #3221: URL: https://github.com/apache/hugegraph/pull/3221#issuecomment-5910090941
Round 4 in 517dfe68: `StorageReadiness.check` without `synchronized`, one in-flight probe shared by concurrent callers with a bounded wait (`StorageReadinessTest` 8/8). The gate on a read through the graph path was built in three variants and measured on k3s: in every one a rolling Store restart takes all Servers out of the Service while the traffic is fine, and with PD down a probe through the shared Store client raises the load failures from 2/244 to 58-87 %, because a failed read invalidates the partition cache the queries use; details in the inline reply and in `results/issue-3212/round4-*` of the validation repo. The gate stays on the ping, its scope is now stated in the Javadoc and in the description; the control battery is 7/7 PASS on this head. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
