Carsten Lohmann
Carsten Lohmann
> we don't do validation on purpose If validation isn't needed here, then my intuition would be that the existing code https://github.com/eclipse-vertx/vert.x/blob/cbf38bfc2cfcceb4ccbae89ba6619f0a5158dff8/src/main/java/io/vertx/core/http/impl/HttpUtils.java#L364-L379 could be adapted by replacing ```` URI uri...
@vietj Any thoughts on the above? Doing the "startsWith("http(s)://" check instead of parsing the URI to check for an absolute URI.
I would also opt for such a request timeout value. Consider this example: ```` vertx.createHttpServer().requestHandler(req -> { logger.info("serverside: got req " + req.path()); if (req.path().contains("slow")) { vertx.setTimer(3000, h -> req.response().end("foo"));...
Right, I misinterpreted the stacktrace here.
I've removed the changes wrapping the MongoDB client operation now. Instead, the existing `mapError` method has been adapted and a new `logError` method has been added.
Log entries corresponding to the recurring errors: ``` Jan 1, 2022 @ 23:38:56.310 WARN org.apache.kafka.clients.producer.internals.Sender [Producer clientId=hono-cmd-router-internal-cmd-sender-uuid] Got error produce response with correlation id 55429 on topic-partition hono.command_internal.hono-adapter-amqp-vertx-66f985899f-hsk85_74894f5112a0_1-0, retrying (2147483646...
I'm still monitoring whether this issue still occurs on a system updated to Hono 1.11 / Kafka 2.8.1. Therefore re-scheduling this for after Hono 1.12.
This issue wasn't observed any more using a Kafka client 2.8.x version including some fixes backported from later Kafka versions. Therefore I think this issue can be closed when doing...
Used Kafka client has been upgraded to 3.2.1, therefore this issue should be obsolete.
In Hono 1.x, using the Hono Helm chart, the liveness/readiness probes were done using HTTPS by default, whereas with Hono 2.0, there was a [switch to plain HTTP](https://github.com/eclipse/packages/commit/ee42aab0542b6ff576394ffbda04a37938b4b20f).