Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[supernova777]

Mounting an AFP Fileserver Share via Mac os 9

Apple menu -> Chooser
select "appleshare"
click "server ip address"
type in the ip address of your server "192.168.x.x"
(in my example my AFP fileserver is a nas4free box running on pc hardware with mirrored/protected data storage)
hit "ok"
a dialog box will appear asing you for your login/pass or whether to simply connect as "guest"
choose the share
the share will appear on your desktop as a "network drive"
and will now be browsable exactly as your local drive, natively within mac os 9 using the familiar file browser

[supernova777]

when i view my files/folders on the afp share with a windows based ftp client it reveals the following hidden files+folders:

.AppleDB
.AppleDesktop
.AppleDouble
.snap
Temporary Items
Network Trash Folder
TheFindByContentFolder
TheVolumeSettingsFolder
.DS_Store
.sujournal

im assuming this is largely to do with how the AFP protocol stashes information regarding the metadata + resource forks etc etc from the macos hfs filesystem

[supernova777]

Mountng an AFP Fileserver share via Mac os X Tiger:

press cmd - K
type in: "afp://192.168.x.x/sharename"
you will get Login/Pass dialog window next
relevant to whatever security settings u set up when u created the fileserver share
(for me this was done via http:// browser configuration tool for nas4free)

the drive now appears as a network share on your desktop

[supernova777]

a list of services supported by my fileserver: (simultaneously)

HAST        
CIFS/SMB        
FTP        
TFTP        
SSH        
NFS        
AFP        
RSYNC        
Unison        
iSCSI Target        
DNLA/UPnP        
iTunes/DAAP        
Dynamic DNS        
SNMP        
UPS        
Webserver        
BitTorrent        
LCDproc

currently i only have CIFS/SMB, FTP + AFP enabled on my box...

[supernova777]

i actually am considering moving my AFP fileserver from a socket 775 pentium 4 box to my oldschool 2002 pentium 3 Tualatin 1.4ghz
because the p3's power requirements are ALMOST ONE THIRD!!! of what the P4 sucks up..while still providing comparible performance!
(84W TDP for the p4 vs 32.2W TDP for the p3!!!)

stats reference:
http://ark.intel.com/products/27525/Intel-Pentium-III-Processor---S-1_40-GHz-512K-Cache-133-MHz-FSB
vs
http://ark.intel.com/products/27465/Intel-Pentium-4-Processor-531-supporting-HT-Technology-1M-Cache-3_00-GHz-800-MHz-SB

more Thermal Design Power figures:
http://en.wikipedia.org/wiki/List_of_CPU_power_dissipation_figures
mac powerpc figures:
http://en.wikipedia.org/wiki/PowerPC_G4
http://macos9lives.com/smforum/index.php?topic=578.0
http://support.apple.com/kb/TA23546?viewlocale=en_US
http://news.cnet.com/Faster-PowerPC-also-saves-energy/2100-1040_3-824621.html


either that or i use a 65W TDP core 2 duo cpu..  it does make sense to use the lowest TDP tho as its going to be on 24/7!

the p4 chip runs hot as hell with a elaborate cooler but the p3 chip is being cooled by the same heatsink that syntho spoke of in his thread about replacing the mdd fans(http://macos9lives.com/smforum/index.php?topic=687.0) the thermaltake volcano



so if u want to run an AFP fileserver on pc hardware to save all your files to frm your macs and gets left on 24/7 this is a real cost effective solution with low electrical cost;)

[supernova777]

the p3 idea is not gonna fly for ZFS:) this filesystem demands more cpu/ram

so i ended up using a e4300 cpu for 1.8ghz core 2 duo.. 8gb of ram...
using 4 x 1TB drives... i set up a RAIDZ1 Array which is giving me 2.6TB of fault-tolerant space
(allows for 1 drive to fail and still the data will be accessible)

so far its working excellent i get speeds of up to 140mb/s writing to the drive, im assuming the read speed is similar..
this is over a gigabit network switch of course

heres some more info on different RAIDZ setups available under nas4free server
https://calomel.org/zfs_raid_speed_capacity.html

 i may bump it up to e8200 because why not, its the same TDP and maybe itll be faster to generate parity data for the raidz1
any other raid5 performance was crap using the legacy UFS filesystem but ZFS is performing quite well with core2duo cpu + lots of ram