[Beowulf] Fundamental HPC cluster problem

John Hearns John.Hearns at xma.co.uk
Tue Jul 21 06:25:19 PDT 2015


http://www.macvendorlookup.com/   shows:

The  eth3 device is manufactured by TP-Link Technologies

The eth0 device by Dell

The eth1 device by Intel

The eth2 device by Intel

I would check the cable to eth3.  Are the lights coming up?
Run ethtool eth3  please

Why are you choosing to use the TP-Link card instead of one of the onboard Intel interfaces?
Your call of course.



From: John Thingstad [mailto:jpthing at online.no]
Sent: 21 July 2015 07:10
To: John Hearns; 'beowulf at beowulf.org'
Subject: Re: [Beowulf] Fundamental HPC cluster problem

john at Pandora:~$ ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: eth3: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast state DOWN group default qlen 1000
    link/ether 14:cc:20:05:f2:1c brd ff:ff:ff:ff:ff:ff
    inet 192.168.10.1/24 brd 192.168.10.255 scope global eth3
       valid_lft forever preferred_lft forever
3: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether d4:ae:52:cc:87:bb brd ff:ff:ff:ff:ff:ff
    inet 192.168.0.100/24 brd 192.168.0.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::d6ae:52ff:fecc:87bb/64 scope link
       valid_lft forever preferred_lft forever
4: eth1: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 90:e2:ba:4c:54:b8 brd ff:ff:ff:ff:ff:ff
5: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 90:e2:ba:4c:54:b9 brd ff:ff:ff:ff:ff:ff

john at Pandora:~$ ip route
default via 192.168.0.1 dev eth0
192.168.0.0/24 dev eth0  proto kernel  scope link  src 192.168.0.100
192.168.10.0/24 dev eth3  proto kernel  scope link  src 192.168.10.1

john at Pandora:~$ sudo service isc-dhcp-server restart
john at Pandora:~$ sudo tail -f /var/log/syslog
...
Jul 21 08:06:13 Pandora kernel: [71959.936774] init: isc-dhcp-server main process (1293) killed by TERM signal
Jul 21 08:06:13 Pandora dhcpd: Not searching LDAP since ldap-server, ldap-port and ldap-base-dn were not specified in the config file
Jul 21 08:06:13 Pandora dhcpd: Not searching LDAP since ldap-server, ldap-port and ldap-base-dn were not specified in the config file
Jul 21 08:06:13 Pandora dhcpd: Wrote 0 leases to leases file.

Den 21. juli 2015 02:49, skrev John Hearns:
John

Please give is the results of ‘ip addr’

Also look closely at the system log messages when dhcp starts.
Stop the dhcp service, then restart it when running a tail –f on the system log.

If that does not help, stop the dhcp service and run dhcpd with the ‘-d’ flag in a separate window,
and give it the argument eth3 on the command line ?



________________________________

Scanned by MailMarshal - M86 Security's comprehensive email content security solution.

________________________________
Any views or opinions presented in this email are solely those of the author and do not necessarily represent those of the company. Employees of XMA Ltd are expressly required not to make defamatory statements and not to infringe or authorise any infringement of copyright or any other legal right by email communications. Any such communication is contrary to company policy and outside the scope of the employment of the individual concerned. The company will not accept any liability in respect of such communication, and the employee responsible will be personally liable for any damages or other liability arising. XMA Limited is registered in England and Wales (registered no. 2051703). Registered Office: Wilford Industrial Estate, Ruddington Lane, Wilford, Nottingham, NG11 7EP
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.beowulf.org/pipermail/beowulf/attachments/20150721/ada88b00/attachment-0001.html>


More information about the Beowulf mailing list