You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Add single destination per block option to benchmark client (#4331)
## Motivation
Right now the default is for every chain to always send transfers to
every other chain, in a way that in each block we cycle through the
chains.
## Proposal
Add an option so that we continue sending chains to every other chain,
but every block contains transactions to just one of the other chains,
and we cycle through chains per block instead of per transfer.
## Test Plan
Deploy a network and test this mode against it.
## Release Plan
- Nothing to do / These changes follow the usual release cycle.
Copy file name to clipboardExpand all lines: CLI.md
+2Lines changed: 2 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -568,6 +568,7 @@ Start a single benchmark process, maintaining a given TPS
568
568
*`--runtime-in-seconds <RUNTIME_IN_SECONDS>` — How long to run the benchmark for. If not provided, the benchmark will run until it is interrupted
569
569
*`--delay-between-chains-ms <DELAY_BETWEEN_CHAINS_MS>` — The delay between chains, in milliseconds. For example, if set to 200ms, the first chain will start, then the second will start 200 ms after the first one, the third 200 ms after the second one, and so on. This is used for slowly ramping up the TPS, so we don't pound the validators with the full TPS all at once
570
570
*`--config-path <CONFIG_PATH>` — Path to YAML file containing chain IDs to send transfers to. If not provided, only transfers between chains in the same wallet
571
+
*`--single-destination-per-block` — Transaction distribution mode. If false (default), distributes transactions evenly across chains within each block. If true, sends all transactions in each block to a single chain, rotating through chains for subsequent blocks
571
572
572
573
573
574
@@ -601,6 +602,7 @@ Run multiple benchmark processes in parallel
601
602
*`--runtime-in-seconds <RUNTIME_IN_SECONDS>` — How long to run the benchmark for. If not provided, the benchmark will run until it is interrupted
602
603
*`--delay-between-chains-ms <DELAY_BETWEEN_CHAINS_MS>` — The delay between chains, in milliseconds. For example, if set to 200ms, the first chain will start, then the second will start 200 ms after the first one, the third 200 ms after the second one, and so on. This is used for slowly ramping up the TPS, so we don't pound the validators with the full TPS all at once
603
604
*`--config-path <CONFIG_PATH>` — Path to YAML file containing chain IDs to send transfers to. If not provided, only transfers between chains in the same wallet
605
+
*`--single-destination-per-block` — Transaction distribution mode. If false (default), distributes transactions evenly across chains within each block. If true, sends all transactions in each block to a single chain, rotating through chains for subsequent blocks
604
606
*`--processes <PROCESSES>` — The number of benchmark processes to run in parallel
0 commit comments