| 1 |
Norman Rieß <norman@×××××××××.org> writes: |
| 2 |
|
| 3 |
>> Is it connected into 10/100 or 1000 (gigabit) setup? |
| 4 |
>> |
| 5 |
|
| 6 |
> It is a gigabit setup. NFS read is about 30-34MB/s, writing is |
| 7 |
> considerably slower with 15MB/s. So writing is a bit slow. But as i do |
| 8 |
> not need fast storage i did not investigate. And it must be mentioned, |
| 9 |
> that the whole data is in AES. |
| 10 |
|
| 11 |
Being AES should have a pretty dramatic impact right? or is it not |
| 12 |
decrypted and just bounced from one place to another? |
| 13 |
|
| 14 |
> I use this share like a local harddisk. There is nothing like "Oh, this |
| 15 |
> is on remote storage, i will do <random thing> differently." I do |
| 16 |
> everything i do on a local disk, and i did not find anything that would |
| 17 |
> not work due to lack of performance. Admitted i do not do much |
| 18 |
> performancecritical stuff. |
| 19 |
|
| 20 |
Thanks for very good input. What you report beats the snot out of the |
| 21 |
WD `My Book World Edition' I'm testing out. I only tried a few tests |
| 22 |
and they weren't done rigorously like someone benchmarking would have |
| 23 |
to do. I made no attempt to control what else might be running, other |
| 24 |
than not purposely starting anything. |
| 25 |
|
| 26 |
I tried copying 950MB of graphic files across gigabit lan (winXP to |
| 27 |
the Book) ... it took 3 min 40 seconds. (about 4mb sec) |
| 28 |
|
| 29 |
Whereas copying the same data from one machine to another (windowsXP) |
| 30 |
took 40 seconds. |
| 31 |
(Incidently.. that appears to be a bit faster than what you report at |
| 32 |
23mb sec) Might have something to do with the fact that it is |
| 33 |
identical filesystem to identical filesystem (ntfs) |
| 34 |
|
| 35 |
Copying the same data from a winXP to my gentoo box across 10/100 lan |
| 36 |
took 1 min 10 seconds. (A little less than what you see at 13mb sec) |
| 37 |
|
| 38 |
So even in a case where the measurement should have been skewed in |
| 39 |
favor of the Book, it was over 300% slower. |
| 40 |
|
| 41 |
And in the case that should have been comparable it was over 500% |
| 42 |
slower. |
| 43 |
|
| 44 |
Unless I've made some horrible error in the math, which is not |
| 45 |
unlikely, I think my test shows 4mb per second. (I just divided the |
| 46 |
MB by the seconds), that is so far under what you see, that alone |
| 47 |
tells me to return this dog and spend the money ($229) building up my |
| 48 |
own. |
Archives