flat assembler
Message board for the users of flat assembler.

Index > Heap > latest bsd os & linux ultimate performance thread.

Author
Thread Post new topic Reply to topic
sleepsleep



Joined: 05 Oct 2006
Posts: 8958
Location: ˛                             ⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣⁣Posts: 334455
sleepsleep
well, this thread would be the most important thread to CONVERT you into other operating system, may it bsd(s) or linux(s).

so, we need to talk numbers here, after all, this is the ultimate figure that let us based on it and make decision.

start performing such tests on your bsd(s) or linux(s) and report us, of course. it is better if your system support dual-boot, so test could be conducted using the same hardware settings.

and most important thing is, this all would be laymen test.

you could do this on windows box too, just please specify
- OS version & filesystem
- memory & processor

try below test.
please quit all other tasks, before performing below tests.
you can use, ur handphone timer to time the process, or any clock will do, precision into seconds is enough.

a1. move 100 mb single file from a location to another location
a2. move 100 mb single file into pendrive/thumbdrive

b1. move 1GB file from a location to another location
b2. move 1GB file from into pendrive/thumbdrive

c1. in localnetwork using ethernet, copy 100 mb and 1GB single file from pcA to pcB
c2. in localnetwork using wifi, copy 100 mb and 1GB single file from pcA to pcB
Post 02 Feb 2009, 02:18
View user's profile Send private message Reply with quote
revolution
When all else fails, read the source


Joined: 24 Aug 2004
Posts: 17340
Location: In your JS exploiting you and your system
revolution
I think any results you get will need to be properly analysed before any conclusions can be made.

Pendrive transfer rates vary greatly depending upon the manufacturer and type of drive controller. Figures from a test here will likely depend almost entirely upon the quality of the pendrive.

HDD transfer rates will similarly be heavily dominated by the HDD interface (IDE/SATA/etc.) plus a further factor is that an OS may cache the source file in memory if it has been recently used. So any results will be dependant upon the users use of the file just before copying and also the HDD's underlying speed.

Network transfer rates will almost certainly be overwhelmingly dominated by the port speed (10M/100M/1G). This speed is considerably slower than other devices in the computer and will most likely be the most significant bottleneck.

Also, since all tests are based upon files then the level of fragmentation of each file may have some impact. Thus a well used file system may show slower results than a new (or recently defragmented) file system even with the same hardware.

In summary, I think you will be trying to compare apples to oranges.


Last edited by revolution on 02 Feb 2009, 04:42; edited 1 time in total
Post 02 Feb 2009, 03:53
View user's profile Send private message Visit poster's website Reply with quote
LocoDelAssembly
Your code has a bug


Joined: 06 May 2005
Posts: 4633
Location: Argentina
LocoDelAssembly
Besides fragmentation, the dual-boot thing will impact the results if every OS has its own source and/or destination file because HDD transfer speed is not the same at all locations (for example, my HDD has a linear read speed of ~92 MB/s at the beginning and ~44 MB at the ending).
Post 02 Feb 2009, 04:12
View user's profile Send private message Reply with quote
drhowarddrfine



Joined: 10 Jul 2007
Posts: 535
drhowarddrfine
You'd have to have the same machines, too. I do know this, FreeBSD can run most Linux apps faster than Linux can due to some architectural reason I don't recall right now.
Post 02 Feb 2009, 04:14
View user's profile Send private message Reply with quote
bitRAKE



Joined: 21 Jul 2003
Posts: 2914
Location: [RSP+8*5]
bitRAKE
I'm not sure if Linux/BSD would support my RAID array - I've always just assumed they would because I have nowhere to backup to. Confused It's more time/cost effective for me to buy redundant disks than backup to another medium. Anyone have experience with this?
Post 02 Feb 2009, 05:15
View user's profile Send private message Visit poster's website Reply with quote
revolution
When all else fails, read the source


Joined: 24 Aug 2004
Posts: 17340
Location: In your JS exploiting you and your system
revolution
bitRAKE: OT - I use a RAID system for the office. But people should realise that RAID is not a good backup solution, it is an availability system. What happens if your RAID controller goes crazy and writes bad data to all the disks in the array? Then you would have no backup. I suggest for your backup you use an independent setup from you main use RAID system.

BTW: The above failure scenario is just one of many types of hardware faults or software bugs that can destroy the entire array.
Post 02 Feb 2009, 06:01
View user's profile Send private message Visit poster's website Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
sleepsleep: you test scenario is pretty flawed and can't really be used for anything serious. Some reasons have already been mentioned by other people - but there's also the "copy file across network" thing... FTP, SMB/CIFS, NFS, SSH/SCP?

Also, simply copying a single file is no good performance indication... you'd need to create some fresh filesystems and do more varied testing. And there's other more interesting tests you haven't touched on, like thread creation time, how fine-grained kernel locking is, scalability etc.

You also haven't specified what usage scenario you want to test - ie., server or desktop use. For desktop users, scalability to 16-core systems is pretty irrelevant, for instance Smile

And last, raw performance isn't going to win anybody over except for pretty specific cases. Other factors (like familarity, documentation and ease-of-use) are going to be a lot more important to most users...
Post 02 Feb 2009, 16:20
View user's profile Send private message Visit poster's website Reply with quote
MichaelH



Joined: 03 May 2005
Posts: 402
MichaelH
Performance testing of different OS doing different tasks should definitely be left to the experts.

However I pose a question that is related, is the XWindows system the achilles heal that will always place any OS using it in the "also run" but got no where category?????? ..... or conversely, is it the best thing since sliced bread and the coming "cloud" nonsense will see XWindows prevail????
Post 03 Feb 2009, 22:39
View user's profile Send private message Reply with quote
Display posts from previous:
Post new topic Reply to topic

Jump to:  


< Last Thread | Next Thread >
Forum Rules:
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum


Copyright © 1999-2020, Tomasz Grysztar. Also on YouTube, Twitter.

Website powered by rwasa.