Benchmark command

[gilbertchen]

This adds a new benchmark command that can be used to measure disk access and network speed.

Usage:

NAME:
   duplicacy benchmark - Run a set of benchmarks to test download and upload speeds

USAGE:
   duplicacy benchmark [command options]  

OPTIONS:
   -file-size <size>		the size of the local file to write to and read from (in MB, default to 256)
   -chunk-count <count>		the number of chunks to upload and download (default to 64)
   -chunk-size <size>		the size of chunks to upload and download (in MB, default to 4)
   -upload-threads <n>		the number of upload threads (default to 1)
   -download-threads <n>	the number of download threads (default to 1)

Sample output:

duplicacy benchmark
Storage set to sftp://[email protected]/storage
Generating 244.14M byte random data in memory
Writing random data to local disk
Wrote 244.14M bytes in 3.05s: 80.00M/s
Reading the random data from local disk
Read 244.14M bytes in 0.18s: 1388.05M/s
Split 244.14M bytes into 53 chunks without compression/encryption in 1.69s: 144.25M/s
Split 244.14M bytes into 53 chunks with compression but without encryption in 2.32s: 105.02M/s
Split 244.14M bytes into 53 chunks with compression and encryption in 2.44s: 99.90M/s
Generating 64 chunks
Uploaded 256.00M bytes in 62.88s: 4.07M/s
Downloaded 256.00M bytes in 63.01s: 4.06M/s
Deleting 64 temporary files

[TheBestPessimist]

Neat idea!

[TheBestPessimist]

Won't the read speed be "wrong" due to caching, since we write the data first to disk, and straight away read it back? I see this even in your example: write = 80MBPS, read = 1388 MBPS (it's over one thousand!).

Second q: why 244 MB for disk operations, and 256 for network?

[leftytennis]

Shouldn't this benchmark command accept a -storage parameter?

[gilbertchen]

I'll merge this PR first and then add an -storage parameter.

@ Yes, it is most likely due to caching but I guess the disk read speed is unlikely the bottleneck so there is really a need to figure out the exact read speed. The difference between 244MB and 2566MB is due to the units being used, which is being addressed by #437.

[gilbertchen]

This pull request has been mentioned on Duplicacy Forum. There might be relevant details there:

http://forum.duplicacy.com/t/benchmark-command-details/1078/1