Added options to benchmark orders vs certs

[oxade]

Reasoning:
The image above roughly shows the execution flow for a FastX transfer request.
Current benchmark measures B-C-D and F-G-H and returns the sum.
From #93, we determined that measuring one or the other might be important.

Solution
This PR provides the option to measure B-C-D and F-G-H separately, or both via CLI selector.
Will enhance this in future to optionally measure end-to-end (B-H and B-L)

Example (outputs are compacted):

Run for OrdersAndCerts: default benchmark of Certs and Orders (B-C-D and F-G-H)

mbp> ./bench --benchmark-type OrdersAndCerts                     
[2021-12-30T16:05:21Z INFO  bench] Starting benchmark of: OrdersAndCerts
.
.
.
[2021-12-30T16:05:29Z INFO  fastpay::network] Done sending Udp requests to 127.0.0.1:9563 (shard 8)
[2021-12-30T16:05:29Z INFO  bench] Received 80000 responses.
[2021-12-30T16:05:29Z WARN  bench] Completed benchmark for OrdersAndCerts
    Total time: 1548267us, items: 40000, tx/sec: 25835.337186673874

Run for OrdersOnly benchmark (B-C-D)

mbp> ./bench --benchmark-type OrdersOnly
[2021-12-30T16:05:33Z INFO  bench] Starting benchmark of: OrdersOnly
[2021-12-30T16:05:33Z INFO  bench] Preparing accounts.
.
.
.
[2021-12-30T16:05:39Z INFO  fastpay::network] Done sending Udp requests to 127.0.0.1:9562 (shard 7)
[2021-12-30T16:05:39Z INFO  bench] Received 40000 responses.
[2021-12-30T16:05:39Z WARN  bench] Completed benchmark for OrdersOnly
    Total time: 463116us, items: 40000, tx/sec: 86371.44905380078

Run for CertsOnly benchmark (F-G-H)

mbp> ./bench --benchmark-type CertsOnly
[2021-12-30T16:05:44Z INFO  bench] Starting benchmark of: CertsOnly
[2021-12-30T16:05:44Z INFO  bench] Preparing accounts.
.
.
.
[2021-12-30T16:05:51Z INFO  fastpay::network] Done sending Udp requests to 127.0.0.1:9567 (shard 12)
[2021-12-30T16:05:51Z INFO  bench] Received 40000 responses.
[2021-12-30T16:05:51Z WARN  bench] Completed benchmark for CertsOnly
   Total time: 1149791us, items: 40000, tx/sec: 34788.931205758265

timeOrders: 463116 + timeCerts: 1149791 = 1612907 is greater than timeOrdersAndCerts: 1548267 as expected

[lxfind]

Cool. Thanks! Could we allow specify these benchmark types by name instead of "benchmark-type 1/2/3"? It will be difficult to remember which one is which when using the command line.

[oxade]

Cool. Thanks! Could we allow specify these benchmark types by name instead of "benchmark-type 1/2/3"? It will be difficult to remember which one is which when using the command line.

@lxfind of course!
Btw if you specify the --help, it tells you what the numbers mean.
However what names do you recommend, if thats easier?

mbp> ./bench --help            
FastPay Benchmark 0.1.0
Local end-to-end test and benchmark of the FastPay protocol
    throughput
......
OPTIONS:
        --benchmark-type <benchmark-type>      Which execution path to track. 1 = Orders + Certs, 2 = Orders Only, 3 =
                                               Certs Only [default: 1]

[lxfind]

I see. If it's not too much code/effort, ideally I think it would be nice to be able to just:

./bench OrdersAndCerts

or

./bench --type OrdersAndCerts

[oxade]

I see. If it's not too much code/effort, ideally I think it would be nice to be able to just:

./bench OrdersAndCerts

or

./bench --type OrdersAndCerts

Done