Increase error threshold for batch length

[andre4i]

286

union Office_Fluid_FluidRuntime_*
| where Data_eventName == "fluid:telemetry:Batching:LengthTooBig" and Data_length >= 1000
| project Event_Time, Data_docId, Data_eventName, Data_hostName, Data_clientType, Data_batchError, Data_length, Data_threshold, Data_duration

There are a lot of instances with unrealistic batch lengths. I believe the problem is using message.sequenceNumber instead of message.clientSequenceNumber when computing the batch length For batches, both numbers have the same semantics. Also, increasing the threshold to 1000.

[vladsud]

There are a lot of instances with unrealistic batch lengths. I believe the problem is using message.sequenceNumber instead of message.clientSequenceNumber when computing the batch length.

There should be no difference. We should add assert for it and not care what we use.