Error calling 'lxd forkstart......

In full details, the exact error message
error: Error calling 'lxd forkstart test-centos-6 /var/lib/lxd/containers /var/log/lxd/test-centos-6/lxc.conf': err='exit status 1'

Again, while rebooting my lapppy after two days, I encountered the above error message again while trying to start my container through LXD. Reading through the LXD issues reports, these are the typical steps to troubleshoot this issue. Note that I've installed the LXD through source code compilation as there are no RPM package available for Fedora 23.

First thing first, as the LXD was built through code compilation, hence it was started manually by running this command. The benefit of starting the LXD daemon this way is that it let you monitor all the debugging messages as shown below.
$ su -c 'lxd --group wheel --debug --verbose'

INFO[11-14|14:10:24] LXD is starting                          path=/var/lib/lxd
WARN[11-14|14:10:24] Per-container AppArmor profiles disabled because of lack of kernel support 
INFO[11-14|14:10:24] Default uid/gid map: 
INFO[11-14|14:10:24]  - u 0 100000 65536 
INFO[11-14|14:10:24]  - g 0 100000 65536 
INFO[11-14|14:10:24] Init                                     driver=storage/dir
INFO[11-14|14:10:24] Looking for existing certificates        cert=/var/lib/lxd/server.crt key=/var/lib/lxd/server.key
DBUG[11-14|14:10:24] Container load                           container=test-busybox
DBUG[11-14|14:10:24] Container load                           container=test-ubuntu-cloud
DBUG[11-14|14:10:24] Container load                           container=test-centos-7
INFO[11-14|14:10:24] LXD isn't socket activated 
INFO[11-14|14:10:24] REST API daemon: 
INFO[11-14|14:10:24]  - binding socket                        socket=/var/lib/lxd/unix.socket

The first step to troubleshoot is to ensure that the default bridge interface, lxcbr0, used by LXD is up and running.
$ ifconfig lxcbr0
lxcbr0: error fetching interface information: Device not found

Next, start the 'lxc-net' service that created this bridge interface. Check if our bridge interface is up.
$ sudo systemctl start lxc-net

$ ifconfig lxcbr0
lxcbr0: flags=4163  mtu 1500
        inet  netmask  broadcast
        inet6 fe80::fcd3:baff:fefd:5bd7  prefixlen 64  scopeid 0x20
        ether fe:7a:fa:dd:06:cd  txqueuelen 0  (Ethernet)
        RX packets 5241  bytes 301898 (294.8 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 7610  bytes 11032257 (10.5 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

Next, check the status of the 'lxc-net' service. Why we need to do so? Remember that the 'lxc-net' service create a virtual switch where three things will be created. First, the bridge itself that links to an existing network interface connecting to the other world. Next, a DNS server which resolves domain name. And lastly, a DHCP server which assigns new IP address to the container. The DNS and DHCP services is provided by the Dnsmasq daemon.
$ sudo systemctl status lxc-net -l

● lxc-net.service - LXC network bridge setup
   Loaded: loaded (/usr/lib/systemd/system/lxc-net.service; enabled; vendor preset: disabled)
   Active: active (exited) since Sat 2015-11-14 16:13:24 MYT; 13s ago
  Process: 9807 ExecStop=/usr/libexec/lxc/lxc-net stop (code=exited, status=0/SUCCESS)
  Process: 9815 ExecStart=/usr/libexec/lxc/lxc-net start (code=exited, status=0/SUCCESS)
 Main PID: 9815 (code=exited, status=0/SUCCESS)
   Memory: 404.0K
      CPU: 46ms
   CGroup: /system.slice/lxc-net.service
           └─9856 dnsmasq -u nobody --strict-order --bind-interfaces --pid-file=/run/lxc/ --listen-address --dhcp-range, --dhcp-lease-max=253 --dhcp-no-override --except-interface=lo --interface=lxcbr0 --dhcp-leasefile=/var/lib/misc/dnsmasq.lxcbr0.leases --dhcp-authoritative

Nov 14 16:13:24 localhost.localdomain dnsmasq[9856]: started, version 2.75 cachesize 150
Nov 14 16:13:24 localhost.localdomain dnsmasq[9856]: compile time options: IPv6 GNU-getopt DBus no-i18n IDN DHCP DHCPv6 no-Lua TFTP no-conntrack ipset auth DNSSEC loop-detect inotify
Nov 14 16:13:24 localhost.localdomain dnsmasq-dhcp[9856]: DHCP, IP range --, lease time 1h
Nov 14 16:13:24 localhost.localdomain dnsmasq-dhcp[9856]: DHCP, sockets bound exclusively to interface lxcbr0
Nov 14 16:13:24 localhost.localdomain dnsmasq[9856]: reading /etc/resolv.conf
Nov 14 16:13:24 localhost.localdomain dnsmasq[9856]: using nameserver
Nov 14 16:13:24 localhost.localdomain dnsmasq[9856]: read /etc/hosts - 2 addresses
Nov 14 16:13:24 localhost.localdomain systemd[1]: Started LXC network bridge setup.

Expect more posts to come on using LXD in Fedora 23.

No comments:

Post a Comment