oops, all MDMF test files were k=3. Need proper results.

[Imported from Trac: page Performance/Sep2011, version 3]
warner 2011-09-26 05:01:56 +00:00
parent c5cb742b59
commit 6b33f4e3dc

@ -22,10 +22,14 @@ significantly with larger k.
## MDMF (trunk) ## MDMF (trunk)
MDMF is fast! Trunk downloads 1MB/10MB/100MB MDMF files at around 4MBps. The MDMF is fast! Trunk downloads 1MB/10MB/100MB MDMF files at around 4MBps.
download speed seems unaffected by k (from 1 to 60). Partial reads take the ~~The download speed seems unaffected by k (from 1 to 60)~~. Partial
expected amount of time: O(data_read), slightly quantized near the 128KiB reads take the expected amount of time: O(data_read), slightly quantized
segment size. near the 128KiB segment size.
(update 25-Sep: the test harness appears to have been flawed, and all
MDMF files were actually k=3. A new test is in the works to get proper
data on how MDMF read speed varies with k).
* MDMF read versus k, 100MB [MDMF-100MB-vs-k.png](../raw/attachments/Performance/Sep2011/MDMF-100MB-vs-k.png) (timing5.out) * MDMF read versus k, 100MB [MDMF-100MB-vs-k.png](../raw/attachments/Performance/Sep2011/MDMF-100MB-vs-k.png) (timing5.out)
* MDMF partial reads, 100MB [MDMF-100MB-partial.png](../raw/attachments/Performance/Sep2011/MDMF-100MB-partial.png) (timing6.out) * MDMF partial reads, 100MB [MDMF-100MB-partial.png](../raw/attachments/Performance/Sep2011/MDMF-100MB-partial.png) (timing6.out)
@ -73,8 +77,8 @@ Likely fixes include:
> alacrity) > alacrity)
* use unencrypted HTTP for share reads * use unencrypted HTTP for share reads
readv() is the least-work/most-promising, since MDMF has readv() and achieves readv() is the least-work/most-promising, ~~since MDMF has readv() and
high speeds. First step is to do whatever MDMF is doing achieves high speeds. First step is to do whatever MDMF is doing~~
## Future Tests ## Future Tests