Subject: Re: [netatalk-admins] Netatalk on a 2 interface Linux box
From: Sascha Knific (knific@ibm.net)
Date: Tue Mar 16 1999 - 21:46:15 EST
Hi Thierry,
I know how frustrating it is to mess around with bad working NICs. I was
doing this with a 3c900 combo for a couple of month - there switches are
nice but the NICs..... (poor Donald Becker - he does all the drivers also
the tulip driver). Do you know how is producing the tulip chip??? Itīs Intel
;---))) - not a joke. I have a P166 running with and Apple (!!!) 100 Mbit
Card (also tulip) - pretty stable.
With this Iīm beating all Win NT Services for Mac and also this extension
for Win NT to do AT over IP. Itīs a great work Adrian does....
Iīm glad I could help you.
Regards
Sascha
At 14:47 15.03.99 -0800, Thierry de Villeneuve wrote:
>Sascha
>
>Thanks for your input on this.
>
>As a follow-up, I can tell you I dumped the eepro100 cards and got them
>replaced by Tulip based NICs (Lynksys 82L169 PNIC). Other friends at
>lthe local Linux user group told me about intrinsic limitation of the
>eepro100 chip and not to rely on it.
>
>Netatalk runs smoothly and my private net doesn't freeze.
>
>Now I'm testing if none of my data get corrupted before installing a
>bigger drive on my netatalk server.
>
>Your help has been crucial to me. Thanks a lot again
>
>Thierry
>
>
>knific@ibm.net wrote:
>>
>> At 12:37 07.03.99 -0800, Thierry de Villeneuve wrote:
>> >Hello. I'm new to the list and haven't find (yet) archives on posters
on the
>> >same subject. I'm sorry to to repost the following.
>> >
>>
>> Hi hi,
>>
>> >Problems:
>> >
>> > 1/ server won't show in chooser, but can be called giving IP/name
>> > 2/ private network freeze, as well as Mac, for few seconds every
>> > few minutes.
>> >
>> >---
>> >
>> >I'm running Netatalk on a multi homed Linux box with RedHat 5.2, kernel
>> >2.0.36, netatalk1.42b_asun2.1.6
>>
>> Whow, you are faster then Adrian. The latest release I got today was pre-
>> asun-2.1.4 ;-)))))))))))
>>
>> >
>> > eth0 goes to my cable modem through DHCP
>> > eth1 goes to my private network - 192.168.1.0
>> >
>> >I'm running named on the box authoritative on my private net.
>> >
>> >AppleVolumes.default contains 2 lines
>> >~ Home
>> >/home/atalk "My share"
>>
>> Thatīs fine...
>>
>> >I type into atalkd.conf only "eth1". Later, it gets rewritten with "phase",
>> >"net" and "addr". There is no way I can set up any zone name as atalkd.conf
>> >docs says.
>> >
>> >How can I set up a Zone name ?
>>
>> Why do you want zones? You need zones to set up routing (AppleTalk ones -
>> not IP)
>> between different networks. Do you want to make your network public? (I
>> donīt think
>> so ;--)))) Since one of the last asun releases we have multiple zones on
>> one interface.
>>
>> I donīt know how to set them up (maybe with nbprgstr - does anybody know?).
>>
>> >---
>> >
>> >When atalk gets started it picks up the hostname from eth0 (the way I
let my
>> >DHCP connection to let go).
>>
>> You can set up the name netatalk uses manually - look in your startup script
>> "atalk" and change the part in witch hostname gets called. (This is one of
>the
>> solutions).
>>
>> >My problem is that no server shows up in the chooser of my Mac on my
private
>> >net.
>> >If I enter the server's address I get the share showing up and I can
mount a
>> >volume.
>>
>> So you have AppleTalk over IP but no EtherTalk...
>>
>> >note: I can indistinctively call the server by either it's name on eth0 or
>> >its name on eth1.
>>
>> Thatīs was named does...
>> Named gets your local name by your definition and the public one by the
>> name server of you ISP (ok, this could be one answer - DNS is complex).
>>
>> >Later, my eth1 network is locked up for a few seconds every minute or
so. IT
>> >freezes the Mac.
>>
>> ????
>> Try the rest of my advises - maybe it goes away ;----)))))))))))
>>
>> Maybe you should concern to make you privat network eth0 (itīs yours and it
>> should be the first on ;----))).
>>
>> >Unmounting any network volume, not shuttong down atalkd won't stop this. I
>> >have to rebooot the gateway/server.
>>
>> Why do you have to restart the server? Do the Macs still freeze after youīve
>> restarted them??
>>
>> Could be a problem with the network card. Check also the remark at the end
>> of this mail.
>>
>> >My /var/log/messages goes like this (see bottom of mail):
>> >
>> >Is there anything a "trained eye" can spot into this that could help me?
>> > - I wonder about the "*" in "dt043n66:AFPServer@*"
>> > - I wonder I may have problem with the fact my box is dual homed and
atalkd
>> >picking up the wrong interface's name.
>>
>> Iīm running a couple of boxes like this but with ISDN and not cable-modems.
>>
>> The message you get is just the info that netatalk registered itself as AFP
>> Server
>> (File Server) with the given name and on no zone ("@*"). The name is not the
>> problem (you can pass any name to netatalk). Maybe netatalk advertises
itself
>> on eth0.
>>
>> >
>> >
>> >That makes a lot of questions.
>> >
>> > Any help, URL, trick will be welcomed.
>>
>> Try this:
>>
>> 1. Set your name manually in script atalk.
>>
>> 2. atalk.conf (created by atalkd):
>> just leave the eth1... line a add "-seed" after "eth1 ". remove eth0 line.
>>
>> 3. afpd.conf
>> ---
>> "BigMac" -transall -address <IP-Address of you Linux box - private side>
>> ---
>>
>> >Thanks,
>> >Thierry
>> >
>> >
>> >
>> >11:54:11 kernel: Appletalk 0.17 for Linux NET3.035
>> >11:54:11 atalkd[473]: restart (1.4b2+asun2.1.2)
>> >11:54:12 atalkd[473]: zip_getnetinfo for eth1
>> >11:54:31 last message repeated 2 times
>> >11:54:41 atalkd[473]: config for no router
>> >11:54:42 atalkd[473]: ready 0/0/0
>> >11:54:54 papd[484]: restart (1.4b2+asun2.1.2)
>> >11:55:01 afpd[493]: dt043n66:AFPServer@* started on 65280.207:128
>> >11:55:01 afpd[493]: ASIP started on 204.210.30.102:548(1) (1.4b2+asun2.1.2)
>> >11:55:50 afpd[494]: ASIP session:548(1) from 192.168.1.50:2067(2)
>> >11:55:50 afpd[493]: server_child[1] 494 done
>> >11:55:54 afpd[495]: ASIP session:548(1) from 192.168.1.50:2068(2)
>> >11:55:54 afpd[495]: login noauth
>> >12:13:30 dhcpcd[157]: got in RENEWING state
>> >12:13:30 dhcpcd[157]: got in BOUND state
>> >12:13:52 kernel: eth1: Transmit timed out: status 0050 command 0000.
>> >12:13:52 kernel: eth1: Trying to restart the transmitter...
>> >12:15:24 kernel: eth1: Transmit timed out: status 0050 command 0000.
>> >12:15:24 kernel: eth1: Trying to restart the transmitter...
>> >12:18:07 kernel: eth1: Transmit timed out: status 0050 command 0000.
>> >12:18:07 kernel: eth1: Trying to restart the transmitter...
>> >12:18:08 kernel: eth1: Transmit timed out: status 0050 command 0000.
>> >12:18:08 kernel: eth1: Trying to restart the transmitter...
>> >12:18:13 kernel: eth1: Transmit timed out: status 0050 command 0000.
>> >12:18:13 kernel: eth1: Trying to restart the transmitter...
>
>
>> Do you have an Intel EtherExpress (Pro)?!!!!!!!!!
>>
>> Check out for a new driver (at cesdis) and(/or???) apply the following
patch:
>> (or us a tulip based card ;--)))))
>> ----------------------------------------------------------------------------
>> --------------------------------
>> On Thu, 25 Feb 1999, Michael H. Martel wrote:
>>
>> > I'm running RedHat 5.2 on an intel Celeron 333 w/Intel EtherExpress Pro
>> > 10/100 card. When I fire up the latest version from a sun I get the
>> > following message :
>> >
>> > eth0: Transmit timed out: Status 0050 Command 0000.
>> > eth0: Trying to restart the transmitter ...
>>
>> Add this line to your /etc/conf.modules file, reboot, and all should be
>> well:
>>
>> options eepro100 multicast_filter_limit=3
>>
>> The 1.0.3 eepro driver that ships with RH 5.2 doesn't support multicast
>> properly.
>>
>> -Jamie
>> ----------------------------------------------------------------------------
>> --------------------------------
>>
>> >12:19:29 afpd[495]: logout
>> >12:19:29 afpd[495]: 1.59KB read, 2.43KB written
>> >12:19:29 afpd[493]: server_child[1] 495 done
>> >12:29:29 kernel: eth1: Transmit timed out: status 0050 command 0000.
>> >12:29:29 kernel: eth1: Trying to restart the transmitter...
>> >12:31:55 afpd[493]: shutting down on signal 15
>> >12:31:55 dt3n66 atalkd[473]: done
>> >
>> >
>> >=== eom =============================================================
>> >Thierry de Villeneuve San Diego, CA 92128
>> >home e-mail thierryv@abac.com voice +1 (619) 679-3366
>> >prof e-mail thierry.de-villeneuve@st.com fax +1 (619) 679-3395
>> >http://users.abac.com/thierryv/ http://www.efsd.org/
>>
>> Good luck...
>>
>> Sascha Knific
This archive was generated by hypermail 2b28 : Sat Dec 18 1999 - 16:16:27 EST