search

Azure / azure-cosmosdb-java

Java Async SDK for SQL API of Azure Cosmos DB
MIT License
54 stars 61 forks source link

The reply message length 4976828 is less than the maximum message length 4194304 #91

Closed zpappa closed 5 years ago

zpappa commented 5 years ago

Description: I realize this isn't the correct forum for this bug - as this is with the MongoDB api not the cosmosdb sql api, but I don't see a related project I can post an issue against for this.

When reading data using Azure Databricks against a Cosmos DB MongoDB API:

df = spark.read.format("com.mongodb.spark.sql.DefaultSource").option("database", "ingress").option("collection", "ingress").option("pipeline", pipeline).load() cnt = df.count()

When trying to perform a find, I receive the following error:

The reply message length 4976828 is less than the maximum message length 4194304

This is completely non-deterministic, sometimes it works sometimes it does not (same pipeline details being provided). The collection being read has roughly 170mb of data over 1mm records. I have 5000 RUs for this collection.,

christopheranderson commented 5 years ago

Sorry for the issue you're facing, but as you pointed out it's not a good forum to discuss this.

Fortunately, we do have a good way to ask questions like this. Please send an email to AskCosmosDB@microsoft.com for issues with MongoDB sutff.

craigjar commented 4 years ago

Was this issue ever resolved? We are facing the same issue, and I can see on Stack Overflow others are as well, but no solution that I can find. Thank you.

RafaelRNunes commented 4 years ago

Was this issue ever resolved? We are facing the same issue, and I can see on Stack Overflow others are as well, but no solution that I can find. Thank you.

I need help too, I can't find a solution yet.