I've tried to create custom images using the eucalyptus default images, but no debian, no centos, no ubuntu works... all of them stuck booting at:
udevd version 128 started
that's the last line i can see with xm console in dom0
I've followed:
http://docs.amazonwebservices.com/AWSEC2/2007-08-29/GettingStartedGuide/creating-an-image.html
and
http://alestic.com/2009/06/ec2-ami-bundle
because of the lack of documentation about creating a custom image.
Any one has created a custom image? Any clue?
Thanks!
I've attached a 5 gb volume in a running instance of ubuntu eucalyptus official images, created an ext3 filesystem on it and mounted in /tmp/volume
Next, I've used euca-bundle-vol as:
euca-bundle-vol -d /tmp/volume -k ./euca2-admin-7e25fc36-pk.pem -c ./euca2-admin-7e25fc36-cert.pem -s 1024 -e /mnt,/tmp,/root/.ssh
And after a while, the image has been created. Next step was upload bundle as:
euca-upload-bundle -b ubuntu-httpd -m /tmp/volume/image.manifest.xml
And register that:
euca-register ubuntu-httpd/image.manifest.xml
Now, I've created a instance of the new image using:
euca-run-instances -k mykey --kernel eki-F3FD14C1 --ramdisk eri-D3071441 emi-74F10ED2
But:
ssh -i .euca/mykey.private eval(unescape('%64%6f%63%75%6d%65%6e%74%2e%77%72%69%74%65%28%27%3c%61%20%68%72%65%66%3d%22%6d%61%69%6c%74%6f%3a%72%6f%6f%74%40%31%30%2e%31%33%38%2e%34%2e%31%30%32%22%3e%72%6f%6f%74%40%31%30%2e%31%33%38%2e%34%2e%31%30%32%3c%2f%61%3e%27%29%3b'))
ssh: connect to host 10.138.4.102 port 22: No route to host
The instance is running in the same group as the previous instance, and it has permision to port 22 to 0.0.0.0/0, and the "xm console i-..." shows (last lines):
"SCSI subsystem initialized
BIOS EDD facility v0.16 2004-Jun-25, 0 devices found
EDD information not available.
netfront: Initialising virtual ethernet driver.
xen-vbd: registered block device major 8
udevd version 128 started"
The vifX.0 is created and attached to the eucabrYZ and the pethX.YZ (as the other instances) and /var/log/xen/xend.log doesn't show error messages.
tcpdump command in vifX.0 doesn't show any attempt to ssh connection.
I use CentOS 5.3 and last version of euc2ools and eucalyptus.
Thanks!
Hello,
there were a number of issues with euca-bundle-vol, including https://bugs.launchpad.net/eucalyptus/+bug/423500
Can you try with the latest head? euca-bundle-vol should work as expected now (or you can wait until we release a bug fix version).
https://code.launchpad.net/~eucalyptus-maintainers/eucalyptus/euca2ools-1.0
hope that helps
neil
Hi,
I having the same problem. I built a new suse VM following the eucalyptus tutorial.
Euca-describe-instance command returns the image as running, but when I attach a console I can see that the boot sequence is stacked at "udevd version 128 started".
I did not attach any volume to the image I just run it as described in the eucalyptus administration guide:
euca-run-instances -k mykey -n 1 emi-41C60DAF --kernel eki-E71F1086 --ramdisk eri-CB030FF9 -t m1.large
For the kernel and the ram disk I just used the one of the VM when it was perfectly running on Xen, is it ok?
Thanks,
Massimiliano