Ask Your Question

smkr's profile - activity

2015-09-17 08:27:39 -0600 commented answer cinder broken iscsiadm: No session found.

@Stify: thanks! I chased the issue over a few good hours: 'tgt-admin -s' was showing 3 targets while on the initiator I only showed 1 target. Stopping the tgtd service and killing those stray processes did the trick! Thanks. I wish I could vote but have no rights yet :)

2015-09-17 01:57:05 -0600 received badge  Supporter (source)
2013-10-08 14:03:47 -0600 answered a question Why does nova-novncproxy blank screen and doesn't connect? - ignoring socket not ready

hi there, this is quite stupid but since I've spent a good hour searching elsewhere: my problem was with using Vimperator Firefox plugin, you must temporarily disable vimperator (<insert> or <s-esc>) then just click on the grey bar and start typing as suggested in the other comments</s-esc></insert>

2013-08-17 11:10:52 -0600 received badge  Famous Question (source)
2013-07-03 08:46:41 -0600 received badge  Notable Question (source)
2013-06-19 14:17:13 -0600 received badge  Popular Question (source)
2013-06-12 10:36:40 -0600 received badge  Editor (source)
2013-06-12 10:35:38 -0600 answered a question python version for grizzly ?

On RedHat 6.4 I use Fedora EPEL.

Grab the RPMs from there and it will ask for all dependencies.

Keep an eye on RDO as suggested by jaybuff.

2013-06-12 06:13:47 -0600 received badge  Nice Answer (source)
2013-06-12 04:02:26 -0600 received badge  Teacher (source)
2013-06-11 11:08:44 -0600 answered a question How to resize a snapshot with Openstack

To shrink a qcow2 image fire up an instance off your snapshot and for each filesystem do:

   dd if=/dev/zero of=fillmeup
   rm fillmeup

f.e. if you've got '/' only:

   cd /
   dd if=/dev/zero of=fillmeup
   sync
   rm -f fillmeup

The same goes for /var, /usr etc.

I've crafter a script for my own usage:

   for q in \
      $(lsblk -ln \
        | grep "^vg_system" \
        | fgrep -v "[SWAP]" \
        | awk '{print $NF}') ; \
   do \
      pushd ${q} ; \
      dd if=/dev/zero of=fillme ; \
      sync ; \
      rm -f fillme ; \
      sync ; \
      popd ; \
   done

From my experience, /dev/zero trick will give you an image about 20 times smaller.

PS. You may be able to avoid booting an instance off the snapshot by using 'guestmount' to mount the snapshot on the hypervisor (tip: use 'virt-df' to display a list of filesystems so you know what you'll be mounting). I haven't tried it (yet) so please reply with your results should you go this route.

2013-06-10 18:58:42 -0600 asked a question nova boot dashboard

Hi,

I can start the VMs from the command line with:

[root@nova ~]# nova boot ostack-vm1 --flavor 3 --image 35f414c9-ad5b-4f95-ba5e-63b8c4a77e60 --nic net-id=3b22fce3-2a00-411e-9706-9c9150a92608 --nic net-id=c6f65c5d-88d5-4984-98a9-6209137ccd55,v4-fixed-ip=192.168.0.201

The 2 networks are fixed IPs pools:

id IPv4 IPv6 start address DNS1 DNS2 VlanID project uuid
1 10.0.0.0/24 None 10.0.0.2 8.8.4.4 None None None 3b22fce3-2a00-411e-9706-9c9150a92608 2 192.168.0.200/29 None 192.168.0.201 192.168.0.1 None None None c6f65c5d-88d5-4984-98a9-6209137ccd55

where the network #2 was created with:

[root@nova ~]# nova-manage network create --label=local-vm --fixed_range_v4=192.168.0.200/29 --bridge=br1 --dns1=192.168.0.1

so the bridge br1 takes care of routing traffic to the "public" network. I don't want Nova to pick a random/available address from the 192.168.0.200/29 pool, that's why the '--nic... v4-fixed-ip=' in the boot command above.

I also run a "management" network on 192.168.1.0/24 that allows me to SSH into the Nova node.

Is there a way to configure Horizon to execute the same command when I create the instance from the Dashboard? It appears that all it does is call 'nova boot' without any networking options. That way I have no control over what 192.168.0.200/29 address is assigned to my instances.

Thanks.