Is there limit for grpc query api?

[ianzhang1988]

I wrote a simple python script to get trace from grpc query api, but I can only get no more than 1000 result, Is there any limit?

my code:

def main():
    channel = grpc.insecure_channel("127.0.0.1:16685")
    stub = query_pb2_grpc.QueryServiceStub(channel)

    time_min = Timestamp()
    time_max = Timestamp()
    time_min.FromDatetime(datetime.fromisoformat("2022-10-11T12:00:00+08:00"))
    time_max.FromDatetime(datetime.fromisoformat("2022-10-12T12:05:00+08:00"))

    print(time_min)
    tqp = query_pb2.TraceQueryParameters(
        service_name="test",
        start_time_min=time_min,
        start_time_max=time_max,
    )

    response = stub.FindTraces(query_pb2.FindTracesRequest(query=tqp))

    counter = 0
    for i in response:
        for ii in i.spans:
            # print(ii)
            # print(ii.trace_id.hex())
            counter+=1
    print(counter)
main()

[yurishkuro]

There is a search_depth you can specify in TraceQueryParameters. Each storage backend defines its own default:

$ grep -rn NumTraces ./plugin/storage | grep 'defaultNumTraces ='
./plugin/storage/cassandra/spanstore/reader.go:67:	defaultNumTraces = 100
./plugin/storage/es/spanstore/reader.go:62:	defaultNumTraces = 100
./plugin/storage/badger/spanstore/reader.go:62:	defaultNumTraces = 100

[ianzhang1988]

A larger value did help, Thanks a lot.

By the way. What search_depth means exactly?

[yurishkuro]

it roughly refers to how much data to return, but the exact interpretation is different between different backends.

[ianzhang1988]

Thanks, very helpful.