Instance unable to retrieve it's medatas during launch (404 errors)
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| OpenStack Compute (nova) |
Invalid
|
Undecided
|
Unassigned | ||
Bug Description
Nova-version :
ii nova-api 2011.2-
ii nova-common 2011.2-
ii nova-compute 2011.2-
ii nova-network 2011.2-
ii nova-objectstore 2011.2-
ii nova-scheduler 2011.2-
ii python-nova 2011.2-
Kernel :
Linux nova-cc1 2.6.35-28-server #50-Ubuntu SMP Fri Mar 18 18:59:25 UTC 2011 x86_64 GNU/Linux
Images repo :
http://
When I run an instance, through euca-run, it's unable to retrieve it's metadatas from the api-server, I've 404 - not found errors
nova-api.log :
http://
euca-get-
http://
iptables (we see that prerouting rule that redirects to the api-server hits that are made to 169.254.169.254):
http://
nova.conf :
http://
When I remove the ec2_dmz_host, I don't have any 404 errors , but "network unreacheable" within the instance.
| Razique Mahroua (razique) wrote | #1 |
| Thierry Carrez (ttx) wrote | #2 |
| Changed in nova: | |
| status: | New → Incomplete |
| Razique Mahroua (razique) wrote | #3 |
| Changed in nova: | |
| status: | Incomplete → Invalid |
I've been able to resolve my issue by doing the following actions :
$ nova-manage project scrub $MY_PROJECT
$ nova-manage network delete 10.0.1.0/27
$ nova-manage network delete 10.0.1.0/26
$ nova-manage network create 10.0.1.0/24 1 256
$ service nova-network restart
=> select * from networks\G;
http://
My idea is that the instance were using a reserved ip by dnsmasq. After I scrub that range and affected a new one to the instance, it is now able to retrieve it's metadatas :
[...]
[ 0.707408] devtmpfs: mounted
[ 0.710454] Freeing unused kernel memory: 880k freed
[ 0.712021] Write protecting the kernel read-only data: 10240k
[ 0.713721] Freeing unused kernel memory: 84k freed
[ 0.722053] Freeing unused kernel memory: 1416k freed
lxcmount stop\/pre-start, process 45
cloud-init start-local running: Wed, 08 Jun 2011 08:27:53 +0000. up 1.93 seconds
no instance data found in start-local
init: cloud-init-local main process (167) terminated with status 1
init: udev-fallback-
init: plymouth main process (47) killed by SEGV signal
init: plymouth-splash main process (277) terminated with status 2
init: cloud-init-nonet main process (228) killed by TERM signal
cloud-init start running: Wed, 08 Jun 2011 08:27:55 +0000. up 3.45 seconds
found data source: DataSourceEc2
[...]