Ryan CrawCour
Ryan CrawCour
to avoid causing a breaking change, could we ignore the clientid if not supplied?
Would you see this metadata being copied in to the actual document that is persisted in Mongo? If so, how would this happen? As a "header" or "schema" complex object?
Stored separately in MongoDB, or elsewhere?
+1 on the Concrete List arg and the Expression to path string and the Fluency. that would make for MUCH neater code. ```csharp IReadOnlyList patchOperations = new[] { PatchOperation.Replace("/ShippedDate", DateTime.UtcNow)...
+1 for the Keep it Simple, and Keep it familiar to other database systems where we can already do this.
Full Connect error log - [2021-01-15 17:18:24,171] ERROR Unexpected exception RntbdTransportClient({"id":1,"isClosed":true,"configuration":{"wireLogLevel":null,"connectionEndpointRediscoveryEnabled":false,"maxChannelsPerEndpoint":130,"maxRequestsPerChannel":30,"bufferPageSize":8192,"maxBufferCapacity":8388608,"requestTimeoutInNanos":5000000000,"requestTimerResolutionInNanos":100000000,"shutdownTimeoutInNanos":15000000000,"connectionAcquisitionTimeoutInNanos":5000000000,"connectTimeoutInMillis":5000,"idleConnectionTimeoutInNanos":0,"idleConnectionTimerResolutionInNanos":100000000,"idleEndpointTimeoutInNanos":3600000000000,"receiveHangDetectionTimeInNanos":65000000000,"sendHangDetectionTimeInNanos":10000000000,"maxConcurrentRequestsPerEndpoint":10000,"userAgent":{"suffix":"APN/1.0 Microsoft/1.0 KafkaConnect/null","userAgent":"azsdk-java-cosmos/4.8.0 Linux/5.4.0-1035-azure JRE/11.0.8 APN/1.0 Microsoft/1.0 KafkaConnect/null"}},"serviceEndpoints":{"count":0,"items":[]}}) is closed received while reading from store. (com.azure.cosmos.implementation.directconnectivity.StoreReader) com.azure.cosmos.implementation.directconnectivity.TransportException: RntbdTransportClient({"id":1,"isClosed":true,"configuration":{"wireLogLevel":null,"connectionEndpointRediscoveryEnabled":false,"maxChannelsPerEndpoint":130,"maxRequestsPerChannel":30,"bufferPageSize":8192,"maxBufferCapacity":8388608,"requestTimeoutInNanos":5000000000,"requestTimerResolutionInNanos":100000000,"shutdownTimeoutInNanos":15000000000,"connectionAcquisitionTimeoutInNanos":5000000000,"connectTimeoutInMillis":5000,"idleConnectionTimeoutInNanos":0,"idleConnectionTimerResolutionInNanos":100000000,"idleEndpointTimeoutInNanos":3600000000000,"receiveHangDetectionTimeInNanos":65000000000,"sendHangDetectionTimeInNanos":10000000000,"maxConcurrentRequestsPerEndpoint":10000,"userAgent":{"suffix":"APN/1.0 Microsoft/1.0 KafkaConnect/null","userAgent":"azsdk-java-cosmos/4.8.0...
it appears to be happening because somewhere a closed comos client instance (that hasn't yet been 'nulled') is trying to be closed, again. possible that this happens when source worker...
can't repro consistently. update Java SDK to latest ver.
splitting conditional logging out in to its own issue #366
> is called there is a network roundtrip, an a load being placed on the backend, it's a request the server has to do work for. **## Agreed Solution** confirmed...