wccpd-devel Mailing List for WCCP Server Daemon

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

Is this lists and project still active ?

Regards,
Ferry

Information requested (plus some other stuff I thought of...

Thanks for the help!

Cheers,
Matt.

[root@clunker root]# lsmod
Module                  Size  Used by    Not tainted
nfs                    76184   2  (autoclean)
lockd                  56016   1  (autoclean) [nfs]
sunrpc                 81052   1  (autoclean) [nfs lockd]
ipt_MARK                1368   1  (autoclean)
ppp_async               9312   1  (autoclean)
ppp_generic            19516   3  (autoclean) [ppp_async]
slhc                    6612   0  (autoclean) [ppp_generic]
wd                      5904   1
ne                      7616   1
isa-pnp                39044   0  [ne]
8390                    7700   0  [wd ne]
crc32                   3728   0  [8390]
af_packet              11048   0  (autoclean)
ipt_state               1048   3  (autoclean)
ipt_LOG                 4184   3  (autoclean)
ipt_TCPMSS              3032   1  (autoclean)
ipt_MASQUERADE          2104   1  (autoclean)
iptable_nat            20888   1  (autoclean) [ipt_MASQUERADE]
ip_conntrack           26376   2  (autoclean) [ipt_state ipt_MASQUERADE 
iptable_nat]
iptable_mangle          2712   1  (autoclean)
iptable_filter          2348   1  (autoclean)
ip_tables              15296  10  [ipt_MARK ipt_state ipt_LOG ipt_TCPMSS 
ipt_MASQUERADE iptable_nat iptable_mangle iptable_filter]
rtc                     8124   0  (autoclean)

[root@clunker root]# service wccpd start
Starting wccpd: [  OK  ]

[root@clunker root]# tail /var/log/messages
Sep 25 21:52:38 clunker wccpd: wccpd startup succeeded
Sep 25 21:52:38 clunker wccpd: WCCP Server daemon started (pid=8306)
Sep 25 21:52:52 clunker wccpd: HERE_I_AM: WebCache at 10.0.0.1 became usable
Sep 25 21:52:52 clunker wccpd: WCCP enabled
Sep 25 21:52:54 clunker kernel: IPVS: Connection hash table configured 
(size=4096, memory=32Kbytes)
Sep 25 21:52:54 clunker kernel: IPVS: ipvs loaded.
Sep 25 21:52:55 clunker kernel: IPVS: [wlc] scheduler registered.
Sep 25 21:52:55 clunker wccpd: WebCache at 10.0.0.1 enabled

[root@clunker root]# lsmod
Module                  Size  Used by    Not tainted
ip_vs_wlc               1572   1  (autoclean)
ip_vs                  67856   2  (autoclean) [ip_vs_wlc]
nfs                    76184   2  (autoclean)
lockd                  56016   1  (autoclean) [nfs]
sunrpc                 81052   1  (autoclean) [nfs lockd]
ipt_MARK                1368   1  (autoclean)
ppp_async               9312   1  (autoclean)
ppp_generic            19516   3  (autoclean) [ppp_async]
slhc                    6612   0  (autoclean) [ppp_generic]
wd                      5904   1
ne                      7616   1
isa-pnp                39044   0  [ne]
8390                    7700   0  [wd ne]
crc32                   3728   0  [8390]
af_packet              11048   0  (autoclean)
ipt_state               1048   3  (autoclean)
ipt_LOG                 4184   3  (autoclean)
ipt_TCPMSS              3032   1  (autoclean)
ipt_MASQUERADE          2104   1  (autoclean)
iptable_nat            20888   1  (autoclean) [ipt_MASQUERADE]
ip_conntrack           26376   2  (autoclean) [ipt_state ipt_MASQUERADE
iptable_nat]
iptable_mangle          2712   1  (autoclean)
iptable_filter          2348   1  (autoclean)
ip_tables              15296  10  [ipt_MARK ipt_state ipt_LOG ipt_TCPMSS
ipt_MASQUERADE iptable_nat iptable_mangle iptable_filter]
rtc                     8124   0  (autoclean)

[root@clunker root]# iptables -vn -L -t mangle
Chain PREROUTING (policy ACCEPT 2481K packets, 956M bytes)
  pkts bytes target     prot opt in     out     source 
destination
  125K   13M MARK       tcp  --  eth0   *      !10.0.0.1 
0.0.0.0/0          tcp dpt:80 MARK set 0x80

[root@clunker root]# ip rule ls
RTNETLINK answers: Invalid argument
Dump terminated

[root@clunker root]# ip route ls
172.31.24.24 dev ppp0  proto kernel  scope link  src 144.137.21.152
10.0.0.0/24 dev eth0  scope link
192.168.1.0/24 dev eth1  scope link
169.254.0.0/16 dev eth1  scope link    <---- I have no idea what this 
is!!! Ooops!
127.0.0.0/8 dev lo  scope link
default dev ppp0  scope link

[root@clunker root]# netstat -rn
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt 
Iface
172.31.24.24    0.0.0.0         255.255.255.255 UH        0 0          0 
ppp0
10.0.0.0        0.0.0.0         255.255.255.0   U         0 0          0 
eth0
192.168.1.0     0.0.0.0         255.255.255.0   U         0 0          0 
eth1
169.254.0.0     0.0.0.0         255.255.0.0     U         0 0          0 
eth1  <--- don't know where this came from!
127.0.0.0       0.0.0.0         255.0.0.0       U         0 0          0 lo
0.0.0.0         0.0.0.0         0.0.0.0         U         0 0          0 
ppp0

> 
> -----Original Message-----
> From: wcc...@li... 
> [mailto:wcc...@li...] On Behalf Of Eduardo J. 
> Blanco
> Sent: Tuesday, 23 September 2003 2:57 AM
> To: Matthew Rex
> Cc: wcc...@li...
> Subject: Re: [wccpd-devel] wccpd and squid talk but still no redirecting
> 
> On Sun, 2003-09-21 at 07:58, Matthew Rex wrote:
> 
>>Hi,
> 
> Hi!
> 
>>I've got a (user) problem with wccpd/ipvs/squid.
>>
>>I have patched and rebuilt a 2.4.22 kernel with ipvs, installed wccpd 
>>from RPM on a linux router and started everything up. I have wccpd and 
>>squid exchanging I See You messages and wccpd reports one cache found 
>>and usable on startup.
>>
>>ipvsadm reports:
>>IP Virtual Server version 1.0.10 (size=4096)
>>Prot LocalAddress:Port Scheduler Flags
>>   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
>>FWM  128 wlc
>>   -> hobbes.localdomain:0         Route   1      0          0
>>
>>I've got iptables matching and marking packets with 0x80 on the linux 
>>router but instead of redirecting to the cache, web traffic passes 
>>through the box as normal. The cache does not see any traffic.
>>
>>Not sure where to look next... do I have to do anything with ipvs? Or 
>>does wccpd take care of all that?
>>
>>I have the client and squid box on the same subnet, a valid setup as I 
>>understand it.
> 
> If squid is on the same subnet as the client, remember to be careful 
> about your firewalling rules to avoid loops. [ie. your squid machine 
> will also use port 80]
> 
>>Can anyone help?
> 
> Maybe your packets are not being routed to localhost (ipvs needs that).  
> wccpd's scripts should take care of that.. but...
> 
> Could you send me:
> - Relevant extracts of syslogd's messages
> - ip rule ls
> - ip route ls
> - iptables -vn -L -t mangle (relevant parts)
> 
> Regards,
> 
> -- 
> Eduardo J. Blanco <ej...@us... 
> <mailto:ej...@us...>>
> 
> 

On Sun, 2003-09-21 at 07:58, Matthew Rex wrote:

> Hi,
> 

Hi!

> I've got a (user) problem with wccpd/ipvs/squid.
> 
> I have patched and rebuilt a 2.4.22 kernel with ipvs, installed wccpd 
> from RPM on a linux router and started everything up. I have wccpd and 
> squid exchanging I See You messages and wccpd reports one cache found 
> and usable on startup.
> 
> ipvsadm reports:
> IP Virtual Server version 1.0.10 (size=4096)
> Prot LocalAddress:Port Scheduler Flags
>    -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
> FWM  128 wlc
>    -> hobbes.localdomain:0         Route   1      0          0
> 
> I've got iptables matching and marking packets with 0x80 on the linux 
> router but instead of redirecting to the cache, web traffic passes 
> through the box as normal. The cache does not see any traffic.
> 
> Not sure where to look next... do I have to do anything with ipvs? Or 
> does wccpd take care of all that?
> 
> I have the client and squid box on the same subnet, a valid setup as I 
> understand it.
> 

If squid is on the same subnet as the client, remember to be careful
about your firewalling rules to avoid loops. [ie. your squid machine
will also use port 80]

> Can anyone help?
> 

Maybe your packets are not being routed to localhost (ipvs needs that). 
wccpd's scripts should take care of that.. but...

Could you send me:
- Relevant extracts of syslogd's messages
- ip rule ls
- ip route ls
- iptables -vn -L -t mangle (relevant parts)

Regards,

-- 
Eduardo J. Blanco <ej...@us...>

Hi,

I've got a (user) problem with wccpd/ipvs/squid.

I have patched and rebuilt a 2.4.22 kernel with ipvs, installed wccpd 
from RPM on a linux router and started everything up. I have wccpd and 
squid exchanging I See You messages and wccpd reports one cache found 
and usable on startup.

ipvsadm reports:
IP Virtual Server version 1.0.10 (size=4096)
Prot LocalAddress:Port Scheduler Flags
   -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
FWM  128 wlc
   -> hobbes.localdomain:0         Route   1      0          0

I've got iptables matching and marking packets with 0x80 on the linux 
router but instead of redirecting to the cache, web traffic passes 
through the box as normal. The cache does not see any traffic.

Not sure where to look next... do I have to do anything with ipvs? Or 
does wccpd take care of all that?

I have the client and squid box on the same subnet, a valid setup as I 
understand it.

Can anyone help?

Thanks,
Matt.

Dear Sir

Many Thanks.i install kernel source as well.i try all the ways.i use this
for compiling
gcc -D__KERNEL__ -I/usr/src/linux-2.2.14/include -Wall -Wstrict-prototypes -
O2 -fomit-frame-pointer \

-fno-strict-
aliasing -pipe -fno-strength-reduce -m386 -DCPU=386 -DMODULE -DMODVERSIONS -
include \

/usr/src/linux-2.2.14/include/linux/modversions.h -c ip_wccp.c

after that it is giving

ip_wccp.c:256 :parse error before '-'
256 :malformed floating constant
3:
20:
22:
25:character constant too long
27:
28:
30:
31:malformed
33:
51:
53:
53:numeric constant constain digits beyond the radix

i use another patch it also giving

ip_wccp.c:256 :parse error
25:constant too long
26:
27:
30:

i try patch p1 < ../ip_wccp as well

plese help me to resolve this issuse

i grately appriciate your help

saman

----- Original Message -----
From: "Eduardo J. Blanco" <ej...@ne...>
To: "saman kumara dissanayake" <sa...@gl...>
Cc: <wcc...@li...>
Sent: Tuesday, February 19, 2002 7:23 PM
Subject: Re: [wccpd-devel] (no subject)

> Hi!
>
> That's weird.  Can you send an strace to me?
>
> Also note that RH 7.0 gcc/glibc has several bugs.
>
> You say your router supports WCCP version 1... now, if you already have
> a router, what are you using wccpd for?
>
> --ejbs
>
> On Tue, 2002-02-19 at 13:59, saman kumara dissanayake wrote:
>     Dear Sir
>
>     I managed to install your wccpd on my linux 7.0 server.but if i
>     restart my server is is starting wccpd no problem.after that if i
>     start squid demon few minutes demon disaper.plese help me out to
>     rectify this problem.i used wccpd ver 1.because of my router support
>     wccp ver 1.
>
>     thanking
>
>

You've got the wrong program if you want to use Squid with a Cisco.

That isn't what wccpd is about, and has nothing to to with running Squid=20
with a Cisco.  You want to patch your kernel's ip_gre module to support=20
WCCP or compile the ip_wccp module.

You can find out about this in the Squid FAQ:

http://www.squid-cache.org/Doc/FAQ/FAQ-17.html

ivan falcon wrote:
> Hi,
>=20
> I'm a novice about this programs , because I want work my squid with a=20
> cisco router
>=20
> The problem is when I try start the wccpd deamon this is the message=20
> that have
> I don't know why
> ./wccpd.sh start
> Starting wccpd: execvp: No existe el fichero o el directorio
>                                                          [FALL=D3]
> thanks
>=20

--=20
Joe Cooper <jo...@sw...>
Web caching appliances and support.
http://www.swelltech.com

Hi,

I'm a novice about this programs , because I want work my squid with a ci=
sco router

The problem is when I try start the wccpd deamon this is the message that=
 have
I don't know why
./wccpd.sh start
Starting wccpd: execvp: No existe el fichero o el directorio
                                                          [FALL=D3]
thanks

--=20
Ivan Falcon
OPERACIONES
COMSAT - PERU
www.comsat.com.pe
telf:  446-9132
iva...@co...

Hi!

That's weird.  Can you send an strace to me?

Also note that RH 7.0 gcc/glibc has several bugs.

You say your router supports WCCP version 1... now, if you already have
a router, what are you using wccpd for?

--ejbs

On Tue, 2002-02-19 at 13:59, saman kumara dissanayake wrote:
    Dear Sir

    I managed to install your wccpd on my linux 7.0 server.but if i
    restart my server is is starting wccpd no problem.after that if i
    start squid demon few minutes demon disaper.plese help me out to
    rectify this problem.i used wccpd ver 1.because of my router support
    wccp ver 1.

    thanking

Dear Sir

I managed to install your wccpd on my linux 7.0 server.but if i restart =
my server is is starting wccpd no problem.after that if i start squid =
demon few minutes demon disaper.plese help me out to rectify this =
problem.i used wccpd ver 1.because of my router support wccp ver 1.

thanking you

saman

On 1 Feb 2002, Eduardo J. Blanco wrote:
> Oops... that's bad.
> Can you attach a tcpdump -lxn -s 2048 udp port 2048 ?
> That would help me figure out what's going on.

well... I've tried to use wccp on another mashine. This time this was=20
Linux. With ip_wccp module loaded (I know it works squid <-> Cisco=20
router).

But still Cisco with ordinary squid does not work together. tcpdump goes=20
priv :) one with --ipip and second one with --gatewaying (dump2 - empty).
When I've tried to add mixed types (dump3) nothing was changed - still no=
t=20
working.

I've got in my logs:
charon  wccpd:            HERE_I_AM: WebCache at 212.1.6.3 became usable
charon  wccpd:            WebCache at 212.1.6.3 enabled
charon  wccpd:            executeCmd: enable-cache: 65280
charon  wccpd:            WebCache at 153.1.0.4 changed hash information
charon  wccpd:            WebCache at 212.1.6.3 changed hash information
charon  wccpd:            WebCache at 153.1.0.4 changed hash information
charon  wccpd:            WebCache at 212.1.6.3 changed hash=20
information

..and so on...

first of all some smarter script must be made, because multiple types of=20
caches with mixed types of encapsulation are unusable not.
Next - when a cache is calling HERE_I_AM wccpd automaticaly enables it.=20
Its wrong, because:
- some cache might be broken
- some cache I might not want to use
- I might to diable this cache from router...

... well another vote for somekind of config :)

>     When you start using something more - you will see some trouble :) =
One=20
>     cache works fine, no matter what kind.
>    =20
>     1. I wanted to configure 1 Cisco cache engine and 2 proxy on squid.=
 I've=20
>     configured first one as encapsulated and the rest as direct routing=
. This=20
>     is what I've became in logs:
>    =20
>     Feb 1 14:17:40 charon  wccpd:            HERE_I_AM: received-id mis=
matched (ours=3D35005211 theirs=3D0)
>     Feb 1 14:18:01 charon  wccpd:            HERE_I_AM: received-id mis=
matched (ours=3D521595368 theirs=3D0)
>     Feb 1 14:18:12 charon  wccpd:            HERE_I_AM: received-id mis=
matched (ours=3D521595368 theirs=3D0)
>     Feb 1 14:18:32 charon  wccpd:            HERE_I_AM: received-id mis=
matched (ours=3D294702567 theirs=3D0)
>     Feb 1 14:18:43 charon  wccpd:            HERE_I_AM: received-id mis=
matched (ours=3D294702567 theirs=3D0)
>     Feb 1 14:19:04 charon  wccpd:            HERE_I_AM: received-id mis=
matched (ours=3D714386896 theirs=3D0)
>    =20
>     And it does not work :( Maybe this is because this one is squid on =
Sun...?=20
>     I don't know - after all current scripts don't allow to make mixed =
types=20
>     of caches (is it possible? should be..)
>    =20
>     2. every manipulation on enable/disable scripts causes that I have =
to=20
>     restart whole wccpd. All of the rules are existenz but no trafic is=
 there.
>    =20
>     3. How is it that wccp after restart remembers last settings? He=20
>     automaticaly starts last servers. Even if I've done _disable_ .=20
>    =20
>     4. Why cache is detected even if protocol isn't correct? Encapsulat=
ion I=20
>     mean. After making restart wccpd detects (restores prom memory?) la=
st=20
>     configuration and when I've changed encapsulation type I have in my=
 logs:
>    =20
>     Feb 1 16:06:25 charon  wccpd:            HERE_I_AM: WebCache at 153=
.1.1.1 became usable
>     Feb 1 16:06:35 charon  wccpd:            WCCP enabled
>     Feb 1 16:06:35 charon  wccpd:            WebCache at 153.1.1.1 enab=
led
>    =20
>     # ipvsadm
>       -> RemoteAddress:Port           Forward Weight ActiveConn InActCo=
nn
>     FWM  128 wlc
>       -> 153.1.1.1:0		  Route   1      230        80
>    =20
>     welll... this is Cisco, so it's wrong.

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

Oops... that's bad.

Can you attach a tcpdump -lxn -s 2048 udp port 2048 ?

That would help me figure out what's going on.

--ejbs

On Fri, 2002-02-01 at 19:05, Blues wrote:

    When you start using something more - you will see some trouble :) One=20
    cache works fine, no matter what kind.
   =20
    1. I wanted to configure 1 Cisco cache engine and 2 proxy on squid. I'v=
e=20
    configured first one as encapsulated and the rest as direct routing. Th=
is=20
    is what I've became in logs:
   =20
    Feb 1 14:17:40 charon  wccpd:            HERE_I_AM: received-id mismatc=
hed (ours=3D35005211 theirs=3D0)
    Feb 1 14:18:01 charon  wccpd:            HERE_I_AM: received-id mismatc=
hed (ours=3D521595368 theirs=3D0)
    Feb 1 14:18:12 charon  wccpd:            HERE_I_AM: received-id mismatc=
hed (ours=3D521595368 theirs=3D0)
    Feb 1 14:18:32 charon  wccpd:            HERE_I_AM: received-id mismatc=
hed (ours=3D294702567 theirs=3D0)
    Feb 1 14:18:43 charon  wccpd:            HERE_I_AM: received-id mismatc=
hed (ours=3D294702567 theirs=3D0)
    Feb 1 14:19:04 charon  wccpd:            HERE_I_AM: received-id mismatc=
hed (ours=3D714386896 theirs=3D0)
   =20
    And it does not work :( Maybe this is because this one is squid on Sun.=
..?=20
    I don't know - after all current scripts don't allow to make mixed type=
s=20
    of caches (is it possible? should be..)
   =20
    2. every manipulation on enable/disable scripts causes that I have to=20
    restart whole wccpd. All of the rules are existenz but no trafic is the=
re.
   =20
    3. How is it that wccp after restart remembers last settings? He=20
    automaticaly starts last servers. Even if I've done _disable_ .=20
   =20
    4. Why cache is detected even if protocol isn't correct? Encapsulation =
I=20
    mean. After making restart wccpd detects (restores prom memory?) last=20
    configuration and when I've changed encapsulation type I have in my log=
s:
   =20
    Feb 1 16:06:25 charon  wccpd:            HERE_I_AM: WebCache at 153.1.1=
.1 became usable
    Feb 1 16:06:35 charon  wccpd:            WCCP enabled
    Feb 1 16:06:35 charon  wccpd:            WebCache at 153.1.1.1 enabled
   =20
    # ipvsadm
      -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
    FWM  128 wlc
      -> 153.1.1.1:0		  Route   1      230        80
   =20
    welll... this is Cisco, so it's wrong.
   =20
    --=20
    ---------------------------------
    pozdr.  Pawe=B3 Go=B3aszewski       =20
    ---------------------------------
    CPU not found - software emulation...
   =20
   =20
   =20
   =20
   =20
   =20
   =20
    _______________________________________________
    wccpd-devel mailing list
    wcc...@li...

https://lists.sourceforge.net/lists/listinfo/wccpd-devel

____________________________________________________
eduardo blanco=20
netlabs
Palmar 2548=20
598 (2)  707 7687=20
ej...@ne...

When you start using something more - you will see some trouble :) One=20
cache works fine, no matter what kind.

1. I wanted to configure 1 Cisco cache engine and 2 proxy on squid. I've=20
configured first one as encapsulated and the rest as direct routing. This=
=20
is what I've became in logs:

Feb 1 14:17:40 charon  wccpd:            HERE_I_AM: received-id mismatche=
d (ours=3D35005211 theirs=3D0)
Feb 1 14:18:01 charon  wccpd:            HERE_I_AM: received-id mismatche=
d (ours=3D521595368 theirs=3D0)
Feb 1 14:18:12 charon  wccpd:            HERE_I_AM: received-id mismatche=
d (ours=3D521595368 theirs=3D0)
Feb 1 14:18:32 charon  wccpd:            HERE_I_AM: received-id mismatche=
d (ours=3D294702567 theirs=3D0)
Feb 1 14:18:43 charon  wccpd:            HERE_I_AM: received-id mismatche=
d (ours=3D294702567 theirs=3D0)
Feb 1 14:19:04 charon  wccpd:            HERE_I_AM: received-id mismatche=
d (ours=3D714386896 theirs=3D0)

And it does not work :( Maybe this is because this one is squid on Sun...=
?=20
I don't know - after all current scripts don't allow to make mixed types=20
of caches (is it possible? should be..)

2. every manipulation on enable/disable scripts causes that I have to=20
restart whole wccpd. All of the rules are existenz but no trafic is there.

3. How is it that wccp after restart remembers last settings? He=20
automaticaly starts last servers. Even if I've done _disable_ .=20

4. Why cache is detected even if protocol isn't correct? Encapsulation I=20
mean. After making restart wccpd detects (restores prom memory?) last=20
configuration and when I've changed encapsulation type I have in my logs:

Feb 1 16:06:25 charon  wccpd:            HERE_I_AM: WebCache at 153.1.1.1=
 became usable
Feb 1 16:06:35 charon  wccpd:            WCCP enabled
Feb 1 16:06:35 charon  wccpd:            WebCache at 153.1.1.1 enabled

# ipvsadm
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
FWM  128 wlc
  -> 153.1.1.1:0		  Route   1      230        80

welll... this is Cisco, so it's wrong.

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

On 30 Jan 2002, Eduardo J. Blanco wrote:
> Don't worry about the flame.

:)

> IMHO, the user shouldn't need to call these scripts directly (because
> wccpd will get confused), and they are not program data either.  That's
> why I put it libexec (from ./configure --help):
> --libexecdir=3DDIR        program executables in DIR [EPREFIX/libexec]
> The way I see it, these scripts are auxiliary executables needed -only-
> by wccpd.

ok - let it be...
It's quite good choice.

>     This 2 thing are connected (in my mind ;)
>     I think that wccpd should call all of the externel programs (ip, ip=
vs).=20
>     Only iptables/ipchains should stay for the end-user.
>     Now - if it will be done (should, I think...):
>     1. one config-file will drive everything. eg. config:
>     cache {
>     	server proxy1.bla.bla.pl;
>     	wccp-protocol 1;
>     	encapsulation gre;
>     	fw-mark 0x80;
>     }
>     2. scripts location will be no problem anymore (no scripts :)  )
>     3. system log will be more reliable - now only connection of new ca=
che is=20
>     noticed. Disconnection isn't. Some diagnostics can be made...? :)
> Sigh... I really don't like this option.  I always wanted to make wccpd
> as portable as possible, that's why I split the OS related stuff from
> the WCCP protocol itself.  You could write similar scripts for BSD, or
> whatever.  Making wccpd call the programs directly only imposes a
> limitation, IMHO.

but... you can make it portable in building stage... Only few functions=20
will differ.=20

You can still keep possibility to execute external scripts ,eg. in config=
:

enable-cache-script:	/bla/bla/script=20
...

You can have flexible and powerfull tool still.

> OTOH, some of the scripts for WCCP v2 will need to call
> ipchains/iptables to cope with dynamic service groups. (ie. services
> described by the webcache dynamically)
> I'm planning to add a config file for WCCPv2, as I have pretty much no
> other option.  I'll take the chance to do something like you suggested.

As you see you don't have choice :)

>     Sure - almost everything can be done by smarter script, but... its =
not=20
>     "clean" solution. Point 3 can't be done in current model.
>    =20
>     Disadvantages of making it that way:
>     - it will be external-programs-dependent
>     - no easy way to change proxies on-fly. Solution of that can be som=
e=20
>     socket or something like that + some small external program for=20
>     management.
>    =20
>     Is it clear?
> About point #3, there are 3 conditions for a cache to 'leave':
> 1) Timeout - presumably the cache is down or rebooted, and it stopped
> sending correct I_AM_HERE messages.
> 2) The designated cache took it out of the assignment hash.
> 3) wccpd is about to exit.
> In all three cases, you should see something like:
> WebCache at xxx.xxx.xxx.xxx disabled
> Let me know if you don't.

It seems I don't.

I've run manualy scripts and no matter how long I left down state - no=20
sign in logs.

> I could change those messages to something like:
> WebCache at xxx.xxx.xxx.xxx expired
> WebCache at xxx.xxx.xxx.xxx disabled by designated WebCache
> WebCache at xxx.xxx.xxx.xxx disabled by wccpd shutdown

This would be cool. More information is better information :)

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

On Tue, 2002-01-29 at 20:27, Blues wrote:
    On 29 Jan 2002, Eduardo J. Blanco wrote:
    >     Few reflections:
    >     - location of scripts - I don't know where should it be, maybe in 
    >     /etc/wccp/scripts ?
    > Sigh... I don't specially like scripts under /etc (yeah, I know,
    > /etc/rc.d).  What does the FSSTD has to say about it?  IIRC, GNU tends
    > to use /usr/libexec for that.

    I'll look at that in  some free time.

    This is strange thing... 
    1. This can be changed to make something work otherwise (esp. other marks)  
    -> /etc/......
    2. like other scripts it can be run-by-hand by root -> /usr/sbin
    3. Maybe standards(TM) are talking something about this kind of files?
    4. I really shure it  shouldn't be in /usr/lib/wccpd - this is not any 
    library.

    ...so... maybe move it to /var/lib/wccpd ? AFAIR FHS says that there 
    should be any of program datas.

    From _user_ point the best thing would be /usr/sbin ....

    Am I making to much flame ? ;)

Don't worry about the flame.

IMHO, the user shouldn't need to call these scripts directly (because
wccpd will get confused), and they are not program data either.  That's
why I put it libexec (from ./configure --help):

--libexecdir=DIR        program executables in DIR [EPREFIX/libexec]

The way I see it, these scripts are auxiliary executables needed -only-
by wccpd.

    >     - dc called in scripts - ugly :)
    > Agreed.  Any way to convert a decimal number to hex or viceversa?  ip
    > rule, iptables and the like all differ slightly on what they expect. :-(
    >     - all caches should be called from some configuration file (maybe in 
    >     /etc/wccp/wccpd.conf.
    > Sorry, I don't understand this... Do you want a config file to list the
    > valid caches?

    This 2 thing are connected (in my mind ;)
    I think that wccpd should call all of the externel programs (ip, ipvs). 
    Only iptables/ipchains should stay for the end-user.
    Now - if it will be done (should, I think...):
    1. one config-file will drive everything. eg. config:
    cache {
    	server proxy1.bla.bla.pl;
    	wccp-protocol 1;
    	encapsulation gre;
    	fw-mark 0x80;
    }
    2. scripts location will be no problem anymore (no scripts :)  )
    3. system log will be more reliable - now only connection of new cache is 
    noticed. Disconnection isn't. Some diagnostics can be made...? :)

Sigh... I really don't like this option.  I always wanted to make wccpd
as portable as possible, that's why I split the OS related stuff from
the WCCP protocol itself.  You could write similar scripts for BSD, or
whatever.  Making wccpd call the programs directly only imposes a
limitation, IMHO.

OTOH, some of the scripts for WCCP v2 will need to call
ipchains/iptables to cope with dynamic service groups. (ie. services
described by the webcache dynamically)

I'm planning to add a config file for WCCPv2, as I have pretty much no
other option.  I'll take the chance to do something like you suggested.

    Sure - almost everything can be done by smarter script, but... its not 
    "clean" solution. Point 3 can't be done in current model.

    Disadvantages of making it that way:
    - it will be external-programs-dependent
    - no easy way to change proxies on-fly. Solution of that can be some 
    socket or something like that + some small external program for 
    management.

    Is it clear?

About point #3, there are 3 conditions for a cache to 'leave':
1) Timeout - presumably the cache is down or rebooted, and it stopped
sending correct I_AM_HERE messages.
2) The designated cache took it out of the assignment hash.
3) wccpd is about to exit.

In all three cases, you should see something like:
WebCache at xxx.xxx.xxx.xxx disabled

Let me know if you don't.

I could change those messages to something like:
WebCache at xxx.xxx.xxx.xxx expired
WebCache at xxx.xxx.xxx.xxx disabled by designated WebCache
WebCache at xxx.xxx.xxx.xxx disabled by wccpd shutdown

About the second problem you describe, I plan to use a default entry for
non-declared caches.  WCCP v2 would, of course, fix most of this issues.

--ejbs

On 29 Jan 2002, Eduardo J. Blanco wrote:
>     Few reflections:
>     - location of scripts - I don't know where should it be, maybe in=20
>     /etc/wccp/scripts ?
> Sigh... I don't specially like scripts under /etc (yeah, I know,
> /etc/rc.d).  What does the FSSTD has to say about it?  IIRC, GNU tends
> to use /usr/libexec for that.

I'll look at that in  some free time.

This is strange thing...=20
1. This can be changed to make something work otherwise (esp. other marks=
) =20
-> /etc/......
2. like other scripts it can be run-by-hand by root -> /usr/sbin
3. Maybe standards(TM) are talking something about this kind of files?
4. I really shure it  shouldn't be in /usr/lib/wccpd - this is not any=20
library.

...so... maybe move it to /var/lib/wccpd ? AFAIR FHS says that there=20
should be any of program datas.

From _user_ point the best thing would be /usr/sbin ....

Am I making to much flame ? ;)

>     - dc called in scripts - ugly :)
> Agreed.  Any way to convert a decimal number to hex or viceversa?  ip
> rule, iptables and the like all differ slightly on what they expect. :-=
(
>     - all caches should be called from some configuration file (maybe i=
n=20
>     /etc/wccp/wccpd.conf.
> Sorry, I don't understand this... Do you want a config file to list the
> valid caches?

This 2 thing are connected (in my mind ;)
I think that wccpd should call all of the externel programs (ip, ipvs).=20
Only iptables/ipchains should stay for the end-user.
Now - if it will be done (should, I think...):
1. one config-file will drive everything. eg. config:
cache {
	server proxy1.bla.bla.pl;
	wccp-protocol 1;
	encapsulation gre;
	fw-mark 0x80;
}
2. scripts location will be no problem anymore (no scripts :)  )
3. system log will be more reliable - now only connection of new cache is=
=20
noticed. Disconnection isn't. Some diagnostics can be made...? :)

Sure - almost everything can be done by smarter script, but... its not=20
"clean" solution. Point 3 can't be done in current model.

Disadvantages of making it that way:
- it will be external-programs-dependent
- no easy way to change proxies on-fly. Solution of that can be some=20
socket or something like that + some small external program for=20
management.

Is it clear?

>     For now it looks OK. It's "raw" but works fine.
>=20
> Ok; I'll see to it.  I'm working now on WCCP v2.=20

I'm waiting :)

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

On Tue, 2002-01-29 at 11:48, Blues wrote:

    On 28 Jan 2002, Eduardo J. Blanco wrote:
    > It's more a hack than a patch.
    > I've read ip_wccp.c, but that is the counterpart of this patch.
    > ip_wccp.c lets you _receive_ WCCP GRE encapsulated packets.  On the
    > other hand, for wccpd I need to load-balance and _send_ WCCP GRE
    > encapsulated packets.
    > 
    > The 'neat' way should be to add a new transport to IPVS, but that would
    > be a much larger patch to both the kernel code and the user-space
    > ipvsadm.  I'd like to try this hack first, as a proof-of-concept.

    This patch works :) Cisco Cache Engine is communicating correctly with 
    wccpd.

    Now I'm making production-testings. For now seems to be OK.

Great!  I'm glad to hear that!

    Few reflections:
    - location of scripts - I don't know where should it be, maybe in 
    /etc/wccp/scripts ?

Sigh... I don't specially like scripts under /etc (yeah, I know,
/etc/rc.d).  What does the FSSTD has to say about it?  IIRC, GNU tends
to use /usr/libexec for that.

    - dc called in scripts - ugly :)

Agreed.  Any way to convert a decimal number to hex or viceversa?  ip
rule, iptables and the like all differ slightly on what they expect. :-(

    - all caches should be called from some configuration file (maybe in 
    /etc/wccp/wccpd.conf.

Sorry, I don't understand this... Do you want a config file to list the
valid caches?

    For now it looks OK. It's "raw" but works fine.

Ok; I'll see to it.  I'm working now on WCCP v2. 

--ejbs

On 28 Jan 2002, Eduardo J. Blanco wrote:
> It's more a hack than a patch.
> I've read ip_wccp.c, but that is the counterpart of this patch.
> ip_wccp.c lets you _receive_ WCCP GRE encapsulated packets.  On the
> other hand, for wccpd I need to load-balance and _send_ WCCP GRE
> encapsulated packets.
>=20
> The 'neat' way should be to add a new transport to IPVS, but that would
> be a much larger patch to both the kernel code and the user-space
> ipvsadm.  I'd like to try this hack first, as a proof-of-concept.

This patch works :) Cisco Cache Engine is communicating correctly with=20
wccpd.

Now I'm making production-testings. For now seems to be OK.

Few reflections:
- location of scripts - I don't know where should it be, maybe in=20
/etc/wccp/scripts ?
- dc called in scripts - ugly :)
- all caches should be called from some configuration file (maybe in=20
/etc/wccp/wccpd.conf.

For now it looks OK. It's "raw" but works fine.

>     I don't like this patch on ipvs... this changes behavior and should=
 add=20
>     new one.
>    =20
>     Take a look at that...
>     http://www.squid-cache.org/WCCP-support/Linux/ip_wccp.c

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

I don't like it either. ;-)

It's more a hack than a patch.

I've read ip_wccp.c, but that is the counterpart of this patch.
ip_wccp.c lets you _receive_ WCCP GRE encapsulated packets.  On the
other hand, for wccpd I need to load-balance and _send_ WCCP GRE
encapsulated packets.

The 'neat' way should be to add a new transport to IPVS, but that would
be a much larger patch to both the kernel code and the user-space
ipvsadm.  I'd like to try this hack first, as a proof-of-concept.

--ejbs

On Sun, 2002-01-27 at 22:22, Blues wrote:

    I don't like this patch on ipvs... this changes behavior and should add=
=20
    new one.
   =20
    Take a look at that...
    http://www.squid-cache.org/WCCP-support/Linux/ip_wccp.c
   =20
    --=20
    ---------------------------------
    pozdr.  Pawe=B3 Go=B3aszewski       =20
    ---------------------------------
    CPU not found - software emulation...
   =20
   =20
    _______________________________________________
    wccpd-devel mailing list
    wcc...@li...

https://lists.sourceforge.net/lists/listinfo/wccpd-devel

____________________________________________________
eduardo blanco=20
netlabs
Palmar 2548=20
598 (2)  707 7687=20
ej...@ne...

I don't like this patch on ipvs... this changes behavior and should add=20
new one.

Take a look at that...
http://www.squid-cache.org/WCCP-support/Linux/ip_wccp.c

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

I'm interested... how many people are using this program. And how many is=
=20
subscibing this list....?

--=20
---------------------------------
pozdr.  Pawe=B3 Go=B3aszewski       =20
---------------------------------
CPU not found - software emulation...

2002	Jan (8)	Feb (5)	Mar	Apr	May (2)	Jun	Jul	Aug	Sep	Oct	Nov	Dec (1)
2003	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep (3)	Oct	Nov	Dec
2006	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug (1)	Sep	Oct	Nov	Dec
2007	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug (2)	Sep	Oct	Nov	Dec

wccpd-devel Mailing List for WCCP Server Daemon

wccpd-devel — Development related discussion.