Problem on running "nova boot myserver"

Thu Aug 2 14:53:38 UTC 2012

On 08/02/2012 02:44 PM, Stephen Liu wrote:
> Hi Derek and all,
> 
> - snip -
> 
> 
>> This is weird, virsh list returned no running vm's but yet something
>> replied to your ssh attempt. I have seen this happen before, when
>> multiple Openstack installations were running on the same local network
>> and shared them same private network. Is there another openstack setup
>> on the same network?
>>
>> this wouldn't answer the question of why the instance failed to start in
>> the first place. can you tail the log files in /var/log/nova/*log
>> to see if any problems show up while you start another instance
> 
> I have another 3 openstacks installed on other VMs running Ubuntu 12.04 as OS.  But they are NOT running at the same time.  I use them for testing ONLY.
> 
> While Fedora 17 is in running they are NOT started.

Ok, requests to 10.0.0.2 are going somewhere, you may need to fire up
traceroute to see where the requests are going, 10.0.0.2 might clash
with the IP range on your local network or across a VPN. If that the
case then you should use another range in the call to
> nova-manage network create
after deleting the old range

<snip/>

> ==> /var/log/nova/cert.log <==
> Command: sh /usr/lib/python2.7/site-packages/nova/CA/genrootca.sh
> Exit code: 1
> Stdout: ''
> Stderr: "cp: cannot create regular file `openssl.cnf': Permission denied\nsed: can't read openssl.cnf: No such file or directory\nerror on line -1 of ./openssl.cnf\n140509211051872:error:02001002:system library:fopen:No such file or directory:bss_file.c:169:fopen('./openssl.cnf','rb')\n140509211051872:error:2006D080:BIO routines:BIO_new_file:no such file:bss_file.c:172:\n140509211051872:error:0E078072:configuration file routines:DEF_LOAD:no such file:conf_def.c:197:\ntouch: cannot touch `index.txt': Permission denied\n/usr/lib/python2.7/site-packages/nova/CA/genrootca.sh: line 28: serial: Permission denied\nUsing configuration from ./openssl.cnf\nerror loading the config file './openssl.cnf'\n140599705130848:error:02001002:system library:fopen:No such file or directory:bss_file.c:169:fopen('./openssl.cnf','rb')\n140599705130848:error:2006D080:BIO routines:BIO_new_file:no such file:bss_file.c:172:\n140599705130848:error:0E078072:configuration file
>  routines:DEF_LOAD:no such file:conf_def.c:197:\n"
> 2012-08-02 21:30:56 AUDIT nova.service [-] Starting cert node (version 2012.1.1-LOCALBRANCH:LOCALREVISION)
> 2012-08-02 21:30:57 CRITICAL nova [-] Unexpected error while running command.
> Command: sh /usr/lib/python2.7/site-packages/nova/CA/genrootca.sh
> Exit code: 1
> Stdout: ''
> Stderr: "cp: cannot create regular file `openssl.cnf': Permission denied\nsed: can't read openssl.cnf: No such file or directory\nerror on line -1 of ./openssl.cnf\n139858879027040:error:02001002:system library:fopen:No such file or directory:bss_file.c:169:fopen('./openssl.cnf','rb')\n139858879027040:error:2006D080:BIO routines:BIO_new_file:no such file:bss_file.c:172:\n139858879027040:error:0E078072:configuration file routines:DEF_LOAD:no such file:conf_def.c:197:\ntouch: cannot touch `index.txt': Permission denied\n/usr/lib/python2.7/site-packages/nova/CA/genrootca.sh: line 28: serial: Permission denied\nUsing configuration from ./openssl.cnf\nerror loading the config file './openssl.cnf'\n140352122591072:error:02001002:system library:fopen:No such file or directory:bss_file.c:169:fopen('./openssl.cnf','rb')\n140352122591072:error:2006D080:BIO routines:BIO_new_file:no such file:bss_file.c:172:\n140352122591072:error:0E078072:configuration file
>  routines:DEF_LOAD:no such file:conf_def.c:197:\n"

I havn't seen this before but the file its trying to create is
/var/lib/nova/CA/openssl.cnf

has the nova user got access to write to this file on your system?
-rw-r--r--. 1 nova nova 2895 Aug  2 15:47 /var/lib/nova/CA/openssl.cnf

infact everything in /var/lib/nova/CA should be owned by nova

<snip/>

> ==> /var/log/nova/scheduler.log <==
> 2012-08-01 17:10:06 INFO nova.rpc.impl_qpid [req-e8d8d756-7751-403c-a348-f8a0f24a63a6 None None] Connected to AMQP server on localhost:5672
> 2012-08-01 19:02:11 AUDIT nova.service [-] Starting scheduler node (version 2012.1.1-LOCALBRANCH:LOCALREVISION)
> 2012-08-01 19:02:12 WARNING nova.db.sqlalchemy.session [req-8b22e296-436c-4bd6-bdcf-1307f511f2b5 None None] SQL connection failed. 10 attempts left.
> 2012-08-01 19:02:23 INFO nova.rpc.impl_qpid [req-8b22e296-436c-4bd6-bdcf-1307f511f2b5 None None] Connected to AMQP server on localhost:5672
> 2012-08-01 23:07:52 AUDIT nova.service [-] Starting scheduler node (version 2012.1.1-LOCALBRANCH:LOCALREVISION)
> 2012-08-01 23:07:55 INFO nova.rpc.impl_qpid [req-21bd76ae-b802-4ed3-bfc3-a76f8903c163 None None] Connected to AMQP server on localhost:5672
> 2012-08-01 23:26:42 WARNING nova.scheduler.manager [req-795fba61-88d9-46a5-babc-c32c5be847b1 02720e417b5b4b7c8f00e1e2e5fc0cf0 59599870461543f9a502a163e3b6931f] Failed to schedule_run_instance: No valid host was found. 
> 2012-08-01 23:26:42 WARNING nova.scheduler.manager [req-795fba61-88d9-46a5-babc-c32c5be847b1 02720e417b5b4b7c8f00e1e2e5fc0cf0 59599870461543f9a502a163e3b6931f] Setting instance d08a6804-3532-4611-81b3-5d8cc8f2a321 to ERROR state.

"Failed to schedule_run_instance: No valid host was found."

this usually means the scheduler couldn't find a suitable compute host
on which to run a VM, an example of why this could happen is that your
compute node doesn't have enough memory to run a VM on. Can you delete
the VM in the error state and try nova boot with --flavor 1 which will
request less memory.

> nova delete <uuid>
> nova boot ....

If memory isn't the problem it may also have something todo with the
error logged in cert.log above