|
aiccu test fails at ping
![[si]](/s/countries/si.gif) Shadow Hawkins on Wednesday, 11 June 2008 10:15:29
Hi!
I have aiccu 20070115-1 on OpenWRT.
Suddenly, "aiccu start" does not create the sixxs net device. This setup worked before.
"aiccu test" fails at first step :
# aiccu test
#######
####### AICCU Quick Connectivity Test
#######
####### [1/8] Ping the IPv4 Local/Your Outer Endpoint (1.2.3.4)
### This should return so called 'echo replies'
### If it doesn't then check your firewall settings
### Your local endpoint should always be pingable
### It could also indicate problems with your IPv4 stack
BusyBox v1.00 (2007.01.30-11:42+0000) multi-call binary
Usage: ping [OPTION]... host
Send ICMP ECHO_REQUEST packets to network hosts.
Options:
-c COUNT Send only COUNT pings.
-s SIZE Send SIZE data bytes in packets (default=56).
-I IPADDR Use IPADDR as source address.
-q Quiet mode, only displays output at start
and when finished.
######
Did this work? [Y/n] n
1.2.3.4 (changed) is the local ppp0 ip.
"ping 1.2.3.4" works fine.
Is aiccu using some strange parameters to ping ?
Because I remember this (eg. "aiicu test") worked before on this system.
Regards,
David
aiccu test fails at ping
Shadow Hawkins on Wednesday, 11 June 2008 10:21:23
Aha, aiccu calls "ping -c 3 -v 1.2.3.4" and ping does not understand the -v option.
aiccu test fails at ping
Shadow Hawkins on Wednesday, 11 June 2008 10:30:38
Just a note: after running "aiccu test" , "aiccu start" works again.
aiccu test fails at ping
Shadow Hawkins on Saturday, 15 September 2012 12:00:42
I just had the same problem again:
- ipv6 not working (ping6 www.google.com gets no response)
- I start "aiccu test" : all works , except IPv4 ping for the reason stated above, the same ping done manually works
- suddenly IPv6 works fine
It seems "aiccu test" does an extra step that is lacking in "aiccu start".
One thing I noticed "ifconfig sixxs" reported "no such device" before "aiccu test", but it appeared afterwards.
Other details:
due to ISP switch I had my tunnel turned off since may 2012. Today I started "aiccu start" , but pinging ipv6 addresses did not work (DNS was resolved, just ping was not working). Then I started "aiccu test" in the same shell and that fixed it.
I'll investigate later, what is the root cause of this...
aiccu test fails at ping
It seems "aiccu test" does an extra step that is lacking in "aiccu start".
The only extra steps that the test does is the tests.
One thing I noticed "ifconfig sixxs" reported "no such device" before "aiccu test", but it appeared afterwards.
Are you using a proto-41 / heartbeat or an AYIYA tunnel? In either case, maybe the IPv6 module or tunnel module (either tun/tap or the proto41 one) is not loaded in time or not available at the time you try to start it.
Today I started "aiccu start" , but pinging ipv6 addresses did not work (DNS was resolved, just ping was not working).
Without outputs/details there is nothing to say about why that does not work.
aiccu test fails at ping
Shadow Hawkins on Monday, 17 September 2012 13:30:16
It is a heartbeat tunnel.
Here are my logs from that day (times are UTC):
Sep 15 11:18:29 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:18:29 (none) local7.info syslog: AICCU running as PID 1032
Sep 15 11:18:31 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:19:00 (none) kern.info kernel: IPv6 v0.8 for NET4.0
Sep 15 11:19:00 (none) kern.debug kernel: vlan0: add 33:33:00:00:00:01 mcast address to master interface
Sep 15 11:19:00 (none) kern.debug kernel: vlan0: add 33:33:ff:xx:xx:xx mcast address to master interface
Sep 15 11:19:00 (none) kern.debug kernel: vlan1: add 33:33:00:00:00:01 mcast address to master interface
Sep 15 11:19:00 (none) kern.debug kernel: vlan1: add 33:33:ff:xx:xx:xx mcast address to master interface
Sep 15 11:19:00 (none) kern.info kernel: IPv6 over IPv4 tunneling driver
Sep 15 11:19:10 (none) kern.debug kernel: eth1: no IPv6 routers present
Sep 15 11:19:10 (none) kern.debug kernel: vlan0: no IPv6 routers present
Sep 15 11:19:10 (none) kern.debug kernel: vlan1: no IPv6 routers present
Sep 15 11:19:10 (none) kern.debug kernel: eth0: no IPv6 routers present
Sep 15 11:19:11 (none) kern.debug kernel: wds0.49153: no IPv6 routers present
Sep 15 11:19:11 (none) kern.debug kernel: wds0.49154: no IPv6 routers present
Sep 15 11:19:11 (none) kern.debug kernel: br0: no IPv6 routers present
Sep 15 11:19:12 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:24:01 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:42:56 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:45:21 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:55:22 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:55:22 (none) local7.info syslog: AICCU running as PID 1240
Meanwhile the router was rebooted and aiccu is not running. Before I start it again I'll wait for any suggestions.
aiccu test fails at ping
Sep 15 11:19:12 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:24:01 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:42:56 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:45:21 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:55:22 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Twice it is detected as already running and then three times it is retrieved? how many times are you starting AICCU?
Likely if you keep starting AICCU that many times you will get blocked from the TIC servers.
aiccu test fails at ping
Shadow Hawkins on Wednesday, 19 September 2012 20:02:28
I started it once, then after I noticed IPv6 is not working I stopped it, tested, start again.
I might have started once or twice, while it was already running.
Normally everything is set up in a few seconds after starting aiccu, that is why I went troubleshooting.
Now I did this:
- ps |grep aiccu # it is not running
- ping6 www.google.com # can`t create raw socket: Address family not supported by protocol
- insmod ipv6
- ping6 www.google.com # no response, DNS is resolved
- aiccu start
- ping6 www.google.com # works
Now IPv6 starte working less than 10 seconds after starting aiccu.
The logs are:
Sep 19 20:01:04 (none) local7.info syslog: Succesfully retrieved tunnel information for T14647
Sep 19 20:01:05 (none) local7.info syslog: AICCU running as PID 1086
Maybe last time the TIC servers were slow to respond ?
aiccu test fails at ping
Shadow Hawkins on Wednesday, 19 September 2012 20:10:28
Now I checked.
"Succesfully retrieved tunnel information for T14647" is printed by "aiccu stop".
So I was not starting it more than one time.
Apparanetly this happened last time:
Sep 15 11:19:12 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:24:01 (none) local7.err syslog: Already running instance HUP'ed, exiting
Sep 15 11:42:56 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:45:21 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Those were from running "aiccu stop" two times. Apparently it got stuck and did not exit immediately. Then I started it again and got the following two lines:
Sep 15 11:55:22 (none) local7.info syslog: Succesfully retrieved tunnel information for Txxxxx
Sep 15 11:55:22 (none) local7.info syslog: AICCU running as PID 1240
Maybe the problem was that I started aiccu too soon after "insmod ipv6" and ipv6 was not ready yet. Who knows...
aiccu test fails at ping
"Succesfully retrieved tunnel information for T14647" is printed by "aiccu stop".
Ah, that is correct. AICCU used to do that. It is really time to release the new version once I have time for it.
Apparently it got stuck and did not exit immediately
One of the many attempts at modifying AICCU could cause that.
Maybe the problem was that I started aiccu too soon after "insmod ipv6" and ipv6 was not ready yet. Who knows...
When the IPv6 module is loaded it is there. Note also that AICCU itself attempts an modprobe ipv6 to make sure that IPv6 is available and it will actually bail out if the stack is not there (based on /proc entries if I recall).
aiccu test fails at ping
Maybe last time the TIC servers were slow to respond ?
The only 'slow' factor can be your connection to them, the boxes themselves run fine.
As you are on OpenWRT though, it might be that you are running modified code which is not behaving like our AICCU. Not much we can tell there.
aiccu test fails at ping
That is because you are using BusyBox which does not support the regular ping options.
|
![[ch]](/s/countries/ch.gif)
on Sunday, 16 September 2012 20:28:18
Posting is only allowed when you are