Re: [LUAU] Cheap linux router

2004-05-24 Thread Myles Uyema
The Linksys routers are targeted toward broadband cable/dsl users.  I've
heard the WRT54G can route a maximum 50Mbps between networks before
reaching peak CPU utilization.  I haven't tested it myself though.

On Wed, 19 May 2004, paul wrote:

Has anyone done any bandwidth testing on these routers?  I would like to know
how many Mbps these can push through them with an empty ruleset.

I have been using a Soekris(http://www.soekris.com) box for my main router.
Not very cheap ($200), but versatile, cool, and quiet.  Most people have been
able to push 17 Mbps through the net4501.  Of course, it's processor is only
an AMD 486.

Always open to new ideas though for small,cool, and quiet routers.  I am
especially looking for one to substain 100Mbps throughput.

Paul

--
Hosted by CyberAddict (http://www.cyberaddict.net)

-- Original Message -------
From: Myles Uyema <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Sent: Wed, 19 May 2004 12:33:06 -0700 (PDT)
Subject: [LUAU] Cheap linux router

> I've been playing around with the Linksys WRT54G and GS models.
> These sport a 200Mhs MIPs cpu and you can flash the firmware to run
> custom Linux builds.
>
> Check out http://docs.sveasoft.com/ and http://openwrt.ksilebo.net/
>
> Linux 2.4.20 kernel, iptables, QoS iproute2+tc stuff all built in.
>
> Horror stories - you can flash the firmware and screw yourself so be
> careful.  These Linksys routers range from $60 - $120 mailorder - retail.
> A really cheap router.  No fans too, and only 12V.
>
> ___
> LUAU@lists.hosef.org mailing list
> http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
--- End of Original Message ---

___
LUAU@lists.hosef.org mailing list
http://lists.hosef.org/cgi-bin/mailman/listinfo/luau



[LUAU] Cheap linux router

2004-05-19 Thread Myles Uyema
I've been playing around with the Linksys WRT54G and GS models.  These
sport a 200Mhs MIPs cpu and you can flash the firmware to run custom Linux
builds.

Check out http://docs.sveasoft.com/ and http://openwrt.ksilebo.net/

Linux 2.4.20 kernel, iptables, QoS iproute2+tc stuff all built in.

Horror stories - you can flash the firmware and screw yourself so be
careful.  These Linksys routers range from $60 - $120 mailorder - retail.
A really cheap router.  No fans too, and only 12V.



Re: [luau] Speaking of hard drives

2003-05-14 Thread Myles Uyema
I know with western digital, you can use your credit card and ask for a 
replacement disk to be shipped to you first.  Then you just use the same 
packaging to return your broken disk.  Your credit card doesn't get 
charged unless you keep both drives =P.

Ask Maxtor if they have something similar.

On Wed, 14 May 2003, stimmy wrote:

I need to RMA a hard drive to Maxtor, but they demand very specific
shipping standards.

So what I'm looking for are two hard drive inserts (top and bottom)
for packaging, that are at least two inches in diameter.

Personally, I think this is ludicrous, for an already broken hard
drive. I mean, Newegg doesn't even ship to these standards.

Anyone have something I could have or purchase at a small fee, in the
Honolulu area?

Thanks!

James

___
LUAU mailing list
[EMAIL PROTECTED]
http://videl.ics.hawaii.edu/mailman/listinfo/luau



[luau] multiplayer starcraft behind firewall

2003-05-11 Thread Myles Uyema
Hey peeps, here's a new script for multiple players with Starcraft 
battle.net behind a linux iptables firewall.

Tested with only 2 computers, so try it out and let me know.
#!/bin/bash
# Written by Myles Uyema; khisanth at uyema d0t net
#
# This is a script to allow Starcraft games to be hosted behind
# a Linux IPTables firewall.
# Tested with Starcraft and 2 machines behind the firewall.
# This may work for other Battle.Net RTS games as well... YMMV

# My Internet IP address
CABLEIP=12.93.33.58

# My PRIVATE LAN Network
# This script assumes Class C network
PRIVLAN=192.168.5

# Battle.Net port usually 6112
BNETPORT=6112

# Enter the last dotted quad IP address of each PC
# We're assuming all the PCs are in a Class C private LAN
# Also, if you have more than 7 PCs, why do you want to get on Battle.net?
# So if my IP address is 192.168.5.5, PC1=5
PC1=5
PC2=98
PC3=
PC4=
PC5=
PC6=
PC7=
PC8=

export CABLEIP PRIVLAN
export PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8

case "$1" in
   start)
  iptables -t nat -F SC-OUT || iptables -t nat -N SC-OUT
  iptables -t nat -F SC-IN || iptables -t nat -N SC-IN

  iptables -t nat -I POSTROUTING -p udp -s ${PRIVLAN}.0/24 --sport 
$BNETPORT -j SC-OUT

  for i in $PC1 $PC2 $PC3 $PC4 $PC5 $PC6 $PC7 $PC8
 do
if [ $i -gt 0 ] ; then
   iptables -t nat -I SC-OUT -s ${PRIVLAN}.${i} -p udp -j SNAT --to 
${CABLEIP}:$((9000+$i))
   iptables -t nat -I PREROUTING -p udp --dport $((9000+$i)) -j 
SC-IN
   iptables -t nat -I SC-IN -d ${CABLEIP} -p udp --dport 
$((9000+$i)) -j DNAT --to ${PRIVLAN}.${i}:${BNETPORT}
fi
 done
  ;;

   stop)
  iptables -t nat -F SC-OUT || exit 0
  iptables -t nat -F SC-IN || exit 0
  iptables -t nat -D POSTROUTING -p udp -s ${PRIVLAN}.0/24 --sport 
$BNETPORT -j SC-OUT

  for i in $PC1 $PC2 $PC3 $PC4 $PC5 $PC6 $PC7 $PC8
 do
if [ $i -gt 0 ] ; then
   iptables -t nat -D PREROUTING -p udp --dport $((9000+$i)) -j 
SC-IN
fi
 done
  iptables -t nat -X SC-OUT
  iptables -t nat -X SC-IN
  ;;

   *)
  echo "Usage: $0 {start|stop}"
  ;;
esac


Re: [luau] Western DIgital Drives / Hard drives in general

2003-01-02 Thread Myles Uyema
Western Digital, IBM-Hitachi, and Maxtor still have 3 year warranties on 
their 'performance' IDE hard drives (8MB cache).

On Thu, 2 Jan 2003, Robert Green wrote:

> Are all industry hard drive warrantees now shortened to 1 year? Are
> any of the drives carrying longer warrantees?



Re: [luau] java swing over ssh

2002-11-22 Thread Myles Uyema
This is just X11 forwarding over ssh.  I'm not well-versed on the X11 
protocol, but I think the 'swing widgets' would need to be somehow 
implemented in the X server.  It's understandable that awt is, since X11 
pretty much built on top of it.


On Fri, 22 Nov 2002, Charles Lockhart wrote:

> Sorry, should have been more clear.  You know how you can "tunnel" an 
> application using ssh, where you ssh to a linux box, and run, say, 
> gedit, and gedit shows up on your desktop, but it's actually running on 
> that other machine and you can open/edit/save files on that other machine.
> 
> Well, you can do the same thing pretty well with java apps written using 
> awt, but for java apps written in swing, performance, even over a fairly 
> good link, just sucks eggs.  I think the awt widgets are supported, but 
> swing widgets aren't, so swing widgets end up being rendered as bitmaps 
> or something.  So what I'm trying to find out is if anyone is working on 
> supporting swing widgets in the same manner that awt widgets are supported.
> 
> Thanks for the links though, I'll check those out.
> 
> -Charles
> 
> Don Brown wrote:
> > Not sure what you mean but there's this:
> > http://www.pitman.co.za/projects/jssh/ and this: http://www.javassh.org
> > and SSH 2.0 here: http://www.jcraft.com/jsch/
> > 
> 
> ___
> LUAU mailing list
> [EMAIL PROTECTED]
> http://videl.ics.hawaii.edu/mailman/listinfo/luau
> 



Re: [luau] NFS ate my raid superblock

2002-11-20 Thread Myles Uyema
Yes, you'll want to watch your kernel messages as you boot up to see if 
it is an FS detection issue or a RAID detection issue.  Having yourself 
on striped RAID-0 though leaves your guard low and your chin up.

Here's the Kernel bootup portion of the RAID log.  It successfully 
detects an MD0 and MD1 device, which are written as special labels on 
each drive/partition that participates in a persistent raid.

md: raid1 personality registered as nr 3
md: Autodetecting RAID arrays.
 [events: 005a]
 [events: 005a]
 [events: 005a]
 [events: 005a]
md: autorun ...
md: considering sdc2 ...
md:  adding sdc2 ...
md:  adding sdb2 ...
md: created md1
md: bind
md: bind
md: running: 
md: sdc2's event counter: 005a
md: sdb2's event counter: 005a
md: RAID level 1 does not need chunksize! Continuing anyway.
md1: max total readahead window set to 508k
md1: 1 data-disks, max readahead per data-disk: 508k
raid1: device sdc2 operational as mirror 1
raid1: device sdb2 operational as mirror 0
raid1: raid set md1 active with 2 out of 2 mirrors
md: updating md1 RAID superblock on device
md: sdc2 [events: 005b]<6>(write) sdc2's sb offset: 17639296
md: sdb2 [events: 005b]<6>(write) sdb2's sb offset: 17639296
md: considering sdc1 ...
md:  adding sdc1 ...
md:  adding sdb1 ...
md: created md0
md: bind
md: bind
md: running: 
md: sdc1's event counter: 005a
md: sdb1's event counter: 005a
md: RAID level 1 does not need chunksize! Continuing anyway.
md0: max total readahead window set to 508k
md0: 1 data-disks, max readahead per data-disk: 508k
raid1: device sdc1 operational as mirror 1
raid1: device sdb1 operational as mirror 0
raid1: raid set md0 active with 2 out of 2 mirrors
md: updating md0 RAID superblock on device
md: sdc1 [events: 005b]<6>(write) sdc1's sb offset: 48064
md: sdb1 [events: 005b]<6>(write) sdb1's sb offset: 48064
md: ... autorun DONE.



On Wed, 20 Nov 2002, Eric Hattemer wrote:

> Wow, thanks a bunch.  Its 1 AM, and I had too much homework to try to fix
> this today, but I'll see if I can work on it tomorrow or later this week.  I
> hope I can remember my block size.
> 
> -Eric Hattemer
> - Original Message -
> From: "Warren Togami" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Wednesday, November 20, 2002 12:45 AM
> Subject: Re: [luau] NFS ate my raid superblock
> 
> 
> > Below is the reply from Gregory Leblanc, maintainer of the Linux RAID
> HOWTO and active in a few other important projects.
> >
> 
> ___
> LUAU mailing list
> [EMAIL PROTECTED]
> http://videl.ics.hawaii.edu/mailman/listinfo/luau
> 



Re: [luau] basic firewall script walkthrough

2002-11-08 Thread Myles Uyema
On Thu, 7 Nov 2002 [EMAIL PROTECTED] wrote:

> So could this be my problem? What would it look like if I had iptables but not
> netfilter installed? iptables at least runs, so I assume I have that 
> installed.

Sounds right.  iptables are the programs.  Netfilter is the kernel 
module portion that iptables requires in order to set up firewall rules.



Re: [luau] Debian and MonMotha's Firewall

2002-10-24 Thread Myles Uyema
You can stick the command into /etc/network/interfaces
(man 5 interfaces to read the manpage)

#Example interfaces file will run firewall.sh after eth0 is brought up
auto eth0
iface eth0 inet dhcp
up /root/firewall.sh

On Thu, 24 Oct 2002, Dean Fujioka wrote:

> I am looking for a place to put MonMotha's firewall script to start 
> automatically at boot. In RedHat, it was rc.local. in SuSE, it was 
> rc.boot. In Debian , I can't find a similar startup script. Is there a 
> Debian demystifier in the house?



Re: [luau] NAT, postfix woes

2002-10-24 Thread Myles Uyema
On Thu, 24 Oct 2002, eXt wrote:

>  > Vague, general questions will receive vague, general answers.
> 
> > 
> > What errors do your PCs get when they discover "no Internet"?
> > Do your PCs get a DNS server or a gateway address through DHCP?
> > Run either "ipconfig /all" or "winipcfg" depending on
> > the Windows OS.
> 
> It gets the "This Page cannot be displayed"
> I had tried the ipconfig /all already, it gives me the machine's IP and then 
> it lists the linux box's (LAN)ip as gateway, DHCP SRVR and DNS SRVR
Okay, then since the LAN network settings look to be okay, then next 
step is to configure the Linux firewall/router to properly forward 
traffic to/from your internal LAN.  Check the "Linux newbie" thread 
which is currently discussing MonMotha's firewall script and quick setup 
of a Linux router.

Someone else is going to have to work through this one.  I don't have 
any postfix experience.  I've only used sendmail.
> 
> > Look in /var/log/* for error messages and try to post some >relevant logs.
> 
> I looked at the log and it just repeats this :
> Oct 23 21:18:02 Server postfix/pickup[28980]: fatal: bad string length (0  
> Look in /var/log/* for error messages and try to post some relevant 
This one looks bad, possibly a misconfiguration.  Stock Redhat Install?

> > logs.
> The Warnings log in /var/mail/  has this on it:
> 
> Oct 20 05:01:03 Server postfix/smtp[32143]: warning: host 
> smtp.latiendita.biz[66.8.154.239] greeted me with my own hostname 
> Server.LaTiendita.biz
> Oct 20 05:01:03 Server postfix/smtp[32143]: warning: host 
> smtp.latiendita.biz[66.8.154.239] replied to HELO/EHLO with my own hostname 
> Server.LaTiendita.biz
> 
> Oct 20 06:58:07 Server postfix/postfix-script: warning: 
> /var/spool/postfix/etc/localtime and /etc/localtime differ
> Oct 20 06:58:07 Server postfix/postfix-script: warning: 
> /var/spool/postfix/etc/hosts and /etc/hosts differ
> Oct 20 06:58:07 Server postfix/postfix-script: warning: 
> /var/spool/postfix/etc/host.conf and /etc/host.conf differ
> 
> Oct 23 21:21:07 Server postfix/master[3424]: warning: process 
> /usr/lib/postfix/smtpd pid 29000 exit status 1
> Oct 23 21:21:07 Server postfix/master[3424]: warning: /usr/lib/postfix/smtpd: 
> bad command startup -- throttling
This one also looks like a misconfiguration, possibly sending bad 
parameters to start the daemon.

> 
> it just repeats those lines



Re: [luau] dummy vim question

2002-10-23 Thread Myles Uyema
Add the following to ~/.vimrc
syntax off

On Tue, 22 Oct 2002, Charles Lockhart wrote:
> I'm running rh7.3.  The default setting for vim is, or seems to be, having
> the syntax highlighting crap turned on (I thought it was supposed to be
> OFF by default).  And I keep looking, but I can't find the damn setup
> files.  I'm about to go postal.  Please keep the free world safe by
> telling me how to turn the damn syntax highlighting crap off forever.
> Please.  I just can't take it anymore.  



Re: [luau] Samba as PDC, yadda yadda yadda

2002-10-22 Thread Myles Uyema
Vague, general questions will receive vague, general answers.

On Tue, 22 Oct 2002, eXt wrote:
> I'm trying to use my Linux box as a PDC for my LAN, but I’m having problems 
> getting samba to even share files.
Samba as a PDC is a fairly advanced step in Samba land.  You have to 
create machine accounts for the actual Windows PCs:  smbpasswd -a -m
Then you have to create user accounts:  smbpasswd -a

I feel Workgroup mode is the way to go in a home setup.
You want to generally create samba user accounts with the same 
name and password as they log into the Windows PC.

Samba user/machine accounts are kept in a separate file from 
/etc/passwd.  Look for the smbpasswd file in /etc/samba/
There are two smbpasswd files.  The program and the user account file.

> I'm also not able to share my WAN connection; my PCs get an IP from the Linux 
> box’s DHCP but no Internet... I tried using the "Server Wizard" included in 
> mandrake 9 but its not working...
What errors do your PCs get when they discover "no Internet"?
Do your PCs get a DNS server or a gateway address through DHCP?
Run either "ipconfig /all" or "winipcfg" depending on the Windows OS.

> And to keep asking questions, I can’t get postfix to work!!! I’m using WEBMIN 
> to try and set it up, but it aunt helping much...
Look in /var/log/* for error messages and try to post some relevant 
logs.

> I need help fast, because every time a user registers on my site the 
> confirmation e-mail is not being sent, so I have to manually e-mail them and 
> make their password, this is a hassle and really unnecessary
Look in /var/log/* for error messages and try to post some relevant 
logs.



Re: [luau] nfs and samba

2002-10-21 Thread Myles Uyema
Quite fine.  Naturally, accessing the same file simultaneously through 
NFS and Samba may bring up some locking issues, so try not to do that.  

I find it generally easier to create windows usernames identical to the 
unix usernames, as it makes user mapping much easier between Unix and 
Windows.

Most likely you'll run Samba in Workgroup mode, and if you create samba 
users (smbpasswd -a) with the same password as their Windows logins, 
it's fairly easy to maintain permissions.

...And hinting at advanced topics, running Samba as a domain 
controller, or possibly in conjunction with an XFS filesystem (Access 
Control Lists for finer grain file permissions).

The www.samba.org FAQ will describe more, and refer you to yet more 
mailing lists =)

> I'm considering sharing a single filesystem using both nfs and samba. Most of
> the discussions I find on the web talk about using samba to re-export a 
> filesystem
> mounted using NFS (and what a bad idea that is). I just want to have an NFS
> daemon for serving unix clients and a smbd for serving windows clients, the
> filesystem itself will be local to both daemons. Any problem here, can samba
> and NFS play nice?