Hyperledger Fabric: Turbocharge Your TPS! | Maximum Throughput (TPS)

[@inzamamansari4007]

That's fantastic, i appreciate your effort. You always help me out with the best content in hyperledger fabric. I have doubt, like we did the test of chaincode performance. But in real world, we must know how much TPS my hyperledger fabric client is giving, and it is very low. I had tested in my case I struggle to get the 5-6 TPS. And single transaction API takes min 2 sec to execute a transaction without waiting till commit

[@adityajoshi12]

This requires enhancement at the chaincode, network and the policy level

[@dyogesh3343]

Found this video really helpful i have used i71400k processor 12 cores to the vm, so my initial tps was 2200, but i later changed to batch size 250 and batchtimeout 5s i got a tps of 6100, which was almost 3 times the initial, my doubt is there any potential negative down side of this if we increase these parameters in the network??

[@adityajoshi12]

Glad this video helped you. The downside of having higher batch size and batch time is that the block commit time will be longer. Lets say hypothetically you have batch size as 1000 and batch time as 60s in that case orderer node will wait to collect either 1000 txns or 60s whichever is earlier and then it will create the block