Ask Your Question

berendt's profile - activity

2014-06-12 14:48:56 -0500 received badge  Teacher (source)
2011-03-17 10:36:11 -0500 answered a question euca-run-instances stuck at scheduling

It would be fine if you could describe your solution here...

2011-03-01 21:02:10 -0500 answered a question NoValidHost: No hosts found

Sorry, should be an information request and no answert...

2011-03-01 21:01:36 -0500 answered a question how to run euca-get-ajax-console ?

Have you set the flag "--ajax_console_proxy_url" in nova.conf? It has to point to your running "nova-ajax-console-proxy" service, for example "http://PROXY_HOST:8000".

And it's not right that you need a "full" linux image, you just have to enable a serial console on the image and I think you have to disable the "write serial console output into a file" entry in the libvirt.xml.template (it's activated per default in the libvirt.xml.template and there are some problems).

Also you should try a newer revision, there were some bugs in nova-ajax-console-proxy, they are fixed in the lastest revision.

2011-03-01 20:54:01 -0500 answered a question NoValidHost: No hosts found

Please post your output of "nova-manage service list". If I correctly remember this means that there are no nova-compute nodes availabe where the nova-scheduler can place a new instance.

2011-02-25 23:34:06 -0500 answered a question KVM images in Bexar

You can use euca-bundle-image, euca-upload-bundle and euca-register to add machine images into your object store. Afterwards try running your image with euca-run-instances. If that's not working for you extract the ramdisk and the kernel from the image and add them to the objectstore, too. Set the flags --ramdisk or --kernel to true and afterwards re-add the machine image and use the flags --ramdisk ID-RAMDISK and --kernel ID-KERNEL. If have not tried the first way, the second one is working for me.

2011-02-25 23:30:45 -0500 answered a question nova-manage --flagfile /etc/nova/nova.conf db sync

Take a look into the wiki: http://wiki.openstack.org/GettingTheCode and try again with the latest revision. Your problem should be solved.

2011-02-25 23:28:39 -0500 answered a question nova-manage project zipfile IRT anne

Have you created the database and synced the database to the latest revision? If not, run "nova-manage db sync" first.

(nova.root): TRACE: OperationalError: (OperationalError) no such table: users u'SELECT users.created_at AS users_created_at, users.updated_at AS users_updated_at, users.deleted_at AS users_deleted_at, users.deleted AS users_deleted, users.id AS users_id, users.name AS users_name, users.access_key AS users_access_key, users.secret_key AS users_secret_key, users.is_admin AS users_is_admin \nFROM users \nWHERE users.id = ? AND users.deleted = ? \n LIMIT 1 OFFSET 0' ('anne', False)

You're missing the tables users. And without this table it should not be possible to create new users, too. Have you successfully created the user "anne" and the project "IRT"? Login to the MySQL database and try "show tables;", "select * from users;" and "select * from projects;". Is this working? Also try a "nova-manage user list" and "nova-manage project list".

2011-02-25 23:22:09 -0500 answered a question Can not access second and next instances via ssh/ping etc

Please paste your nova.conf and your logfiles (I think the logs of nova-network and nova-compute are interesting in this case) to http://paste.openstack.org and post the URLs here.

2011-02-25 12:27:36 -0500 answered a question nova-api --flagfile nova.conf

Which version of python-boto are you using? Which Nova version are you using? Please paste your nova.conf and additional logs to http://paste.openstack.org .

2011-02-23 09:35:54 -0500 answered a question waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id
Metadata forwarding must be handled by the gateway, and since nova does
not do any setup in this mode, it must be done manually.  Requests to
169.254.169.254 port 80 will need to be forwarded to the api server.

There is following method in nova/network/linux_net.py, but it's only called while using Flat DHCP Manager or VLAN Manager.

def metadata_forward(): """Create forwarding rule for metadata""" _confirm_rule("PREROUTING", "-t nat -s 0.0.0.0/0 " "-d 169.254.169.254/32 -p tcp -m tcp --dport 80 -j DNAT " "--to-destination %s:%s" % (FLAGS.ec2_dmz_host, FLAGS.ec2_port))

2011-02-23 09:24:25 -0500 answered a question Installing Nova on a Single Host

That's the wrong boto version, I thing you have to use 1.9 with a patch. Please check the PPA yourself. You can find it here: http://ppa.launchpad.net/openstack/ .

2011-02-23 09:18:20 -0500 answered a question nova-manage project zipfile IRT anne

paste your --sql_connection flag. you created a new database and synced it with nova-manage db sync..?

2011-02-23 09:16:49 -0500 answered a question Please create a database by running a nova-api server on this host.

got to /var/lib/nova/CA and check if there is a file openssl.cnf. If not, copy and modify it from openssl.cnf.tmpl. I'm not sure, but I think you have to call the script genrootca.sh, too.

also we created a symlink /usr/lib64/python/site-packages/CA pointing to /var/lib/nova/CA. I'm not sure if this is necessary or it's only something special about our packaging.

maybe you have no /var/lib/nova/CA and you can find the files directly in /usr/lib64/python/site-packages/CA.

2011-02-22 09:40:35 -0500 answered a question Installing Nova on a Single Host

What version of python-boto are you using? You have to use the one in the PPA. And it's always useful to paste the full error.

2011-02-22 06:56:00 -0500 answered a question How to delete network

I think this has to be possible with bin/nova-manage. I created a bug report here: https://bugs.launchpad.net/nova/+bug/722982 (https://bugs.launchpad.net/nova/+bug/...) .

2011-02-22 06:52:07 -0500 answered a question state hold on pending

fix sudo and try it again (sudo: no tty present and no askpass program specified)...

please use http://paste.openstack.org to paste logfiles, configurations, ... (sorry, forgot the link in my previous posting)

2011-02-22 06:48:25 -0500 answered a question Instance waiting on scheduling

What says the logfile of nova-compute?

2011-02-22 06:47:15 -0500 answered a question Installing Nova on a Single Host

I thought my patch for nova-manage network list were already included in the upstream.. ;)

Login to your database and run "select * from networks;" in your nova-database, this will do the same and you can verify the creation.

you can define the logdir in /etc/nova/nova.conf with "--logdir=/var/log/nova". I think you have to create /var/log/nova manually. Also it's useful to use "--use_syslog" and "--verbose".

have you set the flag "--s3_host=HOST"? Is nova-objectstore running (if not using some other object storage..)? I think euca-upload-bundle does not need a an entry in the networks table, because it's talking with the S3 or Glance and not with the database in any way.

2011-02-22 06:40:35 -0500 answered a question Can't start nova-volume

your error is: sudo: no tty present and no askpass program specified (google it, it's a well known problem)

also you have to create a volume group "nova-volumes" if you want to use nova-volume with the default configuration.

2011-02-21 13:19:46 -0500 answered a question Installing Nova on a Single Host

please paste your /etc/nova/nova.conf and take a look into the logfiles located in /var/log/nova/.

you can create a new network for example with "nova-manage network create 192.168.0.0/24 1 256". Afterwards check if there is a new network with "nova-manage network list".

2011-02-21 13:16:36 -0500 answered a question nova-manage db sync failed

try "nova-manage --flagfile=/etc/nova/nova.conf db sync"

2011-02-20 15:17:02 -0500 answered a question Network Worker crashes on start with Eventlet Error

for users of other distributions:

you have to apply this patch to python-eventlet:

Description: Prevent subprocess module from rewrapping GreenPipes Author: Soren Hansen soren@linux2go.dk Forwarded: https://bitbucket.org/which_linden/eventlet/issue/77/subprocess-module-fails-if-os-module-is (https://bitbucket.org/which_linden/ev...) Last-Update: 2011-02-14

Index: python-eventlet-0.9.14/eventlet/green/subprocess.py

--- python-eventlet-0.9.14.orig/eventlet/green/subprocess.py 2011-02-14 11:31:39.917609581 +0100 +++ python-eventlet-0.9.14/eventlet/green/subprocess.py 2011-02-14 11:31:45.767618158 +0100 @@ -27,7 +27,7 @@ # eventlet.processes.Process.run() method. for attr in "stdin", "stdout", "stderr": pipe = getattr(self, attr) - if pipe is not None: + if pipe is not None and not type(pipe) == greenio.GreenPipe: wrapped_pipe = greenio.GreenPipe(pipe, pipe.mode, bufsize) setattr(self, attr, wrapped_pipe) __init__.__doc__ = subprocess_orig.Popen.__init__.__doc__

2011-02-18 12:19:04 -0500 answered a question bzr branch lp:nova /opt/nova

check the /etc/resolv.conf

2011-02-18 12:15:48 -0500 answered a question state hold on pending

Please append your configuration (/etc/nova/nova.conf), your logfiles (/var/log/nova/*), the used nova version and the used distribution. Without more details you won't get an answer...

2011-02-18 12:13:12 -0500 answered a question Network configuration / Can't ping or ssh to our instances

connect to the running instance through the serial console and look if the network in the instance is coming up.

2011-02-18 12:06:51 -0500 answered a question cp: cannot stat `truncate': No such file or directory

Hmm.. I think it's easier to grep a RPM package from CentOS 6 and to extract the missing binary. I'm using this binary at the moment on a SLES11 SP1 system and it's working fine. /usr/bin/truncate is also not included in the SLES11 coreutils package :(

http://pkgs.org/centos-6-rhel-6/centos-rhel-x86_64/ (http://pkgs.org/centos-6-rhel-6/cento...) http://pkgs.org/centos-6-rhel-6/centos-rhel-x86_64/coreutils-8.4-1.el6.x86_64.rpm.html (http://pkgs.org/centos-6-rhel-6/cento...)