Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
I've checked the obvious things like cables etc, but our DHCP logs show 3 nodes making requests within the last couple of days and they all use the same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the box as a first step.
Matt
Hi Matt, I think you meant: calxeda02-00-03? I get the following when checking for IPs (Matt Gretton-Dann was at my desk asking about Calxeda02 just now :-) ): steve-capper@linaro-gateway:~$ cxmanage ipinfo calxeda02-00-03 Getting IP addresses... 1 successes | 0 errors | 0 nodes left | .
IP info from calxeda02-00-03 Node 0: 192.168.2.75 Node 1: 192.168.2.78 Node 2: 192.168.2.81 Node 3: 192.168.2.84
i.e. only one card active?
So I think it is a good idea to power cycle.
Cheers, -- Steve
On 20 August 2013 10:13, Matt Hart matthew.hart@linaro.org wrote:
Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
I've checked the obvious things like cables etc, but our DHCP logs show 3 nodes making requests within the last couple of days and they all use the same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the box as a first step.
Matt
Ah yes, wrong address, but when scanning for responding IP's I also could only see 4 nodes.
So, power cycled:
matthew-hart@linaro-gateway:~$ cxmanage ipinfo calxeda02-00-03 Getting IP addresses... 1 successes | 0 errors | 0 nodes left | .
IP info from calxeda02-00-03 Node 0: 192.168.2.75 Node 1: 192.168.2.78 Node 2: 192.168.2.81 Node 3: 192.168.2.84 Node 4: 192.168.2.87 Node 5: 192.168.2.90 Node 6: 192.168.2.93 Node 7: 192.168.2.96 Node 8: 192.168.2.99 Node 9: 192.168.2.102 Node 10: 192.168.2.105 Node 11: 192.168.2.108 Node 12: 192.168.2.111 Node 13: 192.168.2.114 Node 14: 192.168.2.117 Node 15: 192.168.2.120 Node 16: 192.168.2.123 Node 17: 192.168.2.126 Node 18: 192.168.2.129 Node 19: 192.168.2.132 Node 20: 192.168.2.135 Node 21: 192.168.2.138 Node 22: 192.168.2.141 Node 23: 192.168.2.144
Much better.
Thanks all, Matt.
On 20 August 2013 10:24, Steve Capper steve.capper@linaro.org wrote:
Hi Matt, I think you meant: calxeda02-00-03? I get the following when checking for IPs (Matt Gretton-Dann was at my desk asking about Calxeda02 just now :-) ): steve-capper@linaro-gateway:~$ cxmanage ipinfo calxeda02-00-03 Getting IP addresses... 1 successes | 0 errors | 0 nodes left | .
IP info from calxeda02-00-03 Node 0: 192.168.2.75 Node 1: 192.168.2.78 Node 2: 192.168.2.81 Node 3: 192.168.2.84
i.e. only one card active?
So I think it is a good idea to power cycle.
Cheers,
Steve
On 20 August 2013 10:13, Matt Hart matthew.hart@linaro.org wrote:
Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
I've checked the obvious things like cables etc, but our DHCP logs show 3 nodes making requests within the last couple of days and they all use the same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the
box
as a first step.
Matt
Steve Capper steve.capper@linaro.org writes:
Hi Matt, I think you meant: calxeda02-00-03?
I'm confused now. AIUI (and linaro-gateway:/etc/dnsmasq.conf backs me up on this) the naming scheme is as follows:
calxeda{chassis_number}-{node_number}-{interface_number}
where {interface_number} is:
0 -- eth0 in the node 1 -- eth1 in the node 2 -- ecme for the node
...
Oh! There are entries for calxeda01-00-03 and calxeda01-00-03 in /etc/hosts. Who did that?
Cheers, mwh
I get the following when checking for IPs (Matt Gretton-Dann was at my desk asking about Calxeda02 just now :-) ): steve-capper@linaro-gateway:~$ cxmanage ipinfo calxeda02-00-03 Getting IP addresses... 1 successes | 0 errors | 0 nodes left | .
IP info from calxeda02-00-03 Node 0: 192.168.2.75 Node 1: 192.168.2.78 Node 2: 192.168.2.81 Node 3: 192.168.2.84
i.e. only one card active?
So I think it is a good idea to power cycle.
Cheers,
Steve
On 20 August 2013 10:13, Matt Hart matthew.hart@linaro.org wrote:
Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
I've checked the obvious things like cables etc, but our DHCP logs show 3 nodes making requests within the last couple of days and they all use the same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the box as a first step.
Matt
linaro-validation mailing list linaro-validation@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-validation
Pass on that one, no idea as it was probably done before my time!
As for the DHCP problem, it's worth noting that we aren't running the latest firmware so I'm due to schedule some downtime for the Calxedas to upgrade them. To backup what you've reported, our problems started when I did some maintenance on the lab on Saturday and restarted dnsmasq a few times. No other devices in the lab had trouble with this but 20 of the 24 nodes on the Calxeda stopped renewing their DHCP lease.
Cheers, Matt
On 20 August 2013 22:41, Michael Hudson-Doyle michael.hudson@linaro.orgwrote:
Steve Capper steve.capper@linaro.org writes:
Hi Matt, I think you meant: calxeda02-00-03?
I'm confused now. AIUI (and linaro-gateway:/etc/dnsmasq.conf backs me up on this) the naming scheme is as follows:
calxeda{chassis_number}-{node_number}-{interface_number}
where {interface_number} is:
0 -- eth0 in the node 1 -- eth1 in the node 2 -- ecme for the node
...
Oh! There are entries for calxeda01-00-03 and calxeda01-00-03 in /etc/hosts. Who did that?
Cheers, mwh
I get the following when checking for IPs (Matt Gretton-Dann was at my desk asking about Calxeda02 just now :-) ): steve-capper@linaro-gateway:~$ cxmanage ipinfo calxeda02-00-03 Getting IP addresses... 1 successes | 0 errors | 0 nodes left | .
IP info from calxeda02-00-03 Node 0: 192.168.2.75 Node 1: 192.168.2.78 Node 2: 192.168.2.81 Node 3: 192.168.2.84
i.e. only one card active?
So I think it is a good idea to power cycle.
Cheers,
Steve
On 20 August 2013 10:13, Matt Hart matthew.hart@linaro.org wrote:
Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
I've checked the obvious things like cables etc, but our DHCP logs show
3
nodes making requests within the last couple of days and they all use
the
same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the
box
as a first step.
Matt
linaro-validation mailing list linaro-validation@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-validation
On 20 Aug 2013, at 22:41, Michael Hudson-Doyle michael.hudson@linaro.org wrote:
Steve Capper steve.capper@linaro.org writes:
Hi Matt, I think you meant: calxeda02-00-03?
I'm confused now. AIUI (and linaro-gateway:/etc/dnsmasq.conf backs me up on this) the naming scheme is as follows:
calxeda{chassis_number}-{node_number}-{interface_number}
where {interface_number} is:
0 -- eth0 in the node 1 -- eth1 in the node 2 -- ecme for the node
...
Oh! There are entries for calxeda01-00-03 and calxeda01-00-03 in /etc/hosts. Who did that?
That was me when we were first deploying the Calxedas, which was coincident with me deploying dnsmasq.
Dave
Cheers, mwh
I get the following when checking for IPs (Matt Gretton-Dann was at my desk asking about Calxeda02 just now :-) ): steve-capper@linaro-gateway:~$ cxmanage ipinfo calxeda02-00-03 Getting IP addresses... 1 successes | 0 errors | 0 nodes left | .
IP info from calxeda02-00-03 Node 0: 192.168.2.75 Node 1: 192.168.2.78 Node 2: 192.168.2.81 Node 3: 192.168.2.84
i.e. only one card active?
So I think it is a good idea to power cycle.
Cheers,
Steve
On 20 August 2013 10:13, Matt Hart matthew.hart@linaro.org wrote:
Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
I've checked the obvious things like cables etc, but our DHCP logs show 3 nodes making requests within the last couple of days and they all use the same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the box as a first step.
Matt
linaro-validation mailing list linaro-validation@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-validation
linaro-validation mailing list linaro-validation@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-validation
Matt Hart matthew.hart@linaro.org writes:
Hi all,
For reasons I'm yet to establish, the management node of calxeda02 has become unresponsive and unreachable on our network.
linaro-gateway:~$ cxmanage macaddrs calxeda02-00-02 Getting MAC addresses... 0 successes | 1 errors | 0 nodes left | .
Command failed on these hosts calxeda02-00-02 : [Errno -2] Name or service not known
So I've been observing behaviour where the ECME nodes stop renewing their DHCP leases, which means that dnsmasq stops resolving the name associated with them. Using the IP address works in this case.
There's a bug from me in the calxeda tracker about this.
I've checked the obvious things like cables etc, but our DHCP logs show 3 nodes making requests within the last couple of days and they all use the same link cable so it doesn't seem a physical networking issue.
If nobody has any reservations or comments, I'd like to power cycle the box as a first step.
That will fix the problem for a few days/weeks if it is what I suspected :)
Cheers, mwh
linaro-validation@lists.linaro.org
-
Dave Pigott -
Matt Hart -
Michael Hudson-Doyle -
Steve Capper