Ask Your Question

likid0's profile - activity

2019-02-08 05:34:44 -0500 received badge  Famous Question (source)
2019-02-08 05:34:44 -0500 received badge  Notable Question (source)
2019-02-08 05:34:44 -0500 received badge  Popular Question (source)
2015-04-15 23:43:49 -0500 received badge  Famous Question (source)
2015-04-01 12:05:37 -0500 received badge  Notable Question (source)
2015-03-31 12:32:47 -0500 received badge  Student (source)
2015-03-31 11:19:36 -0500 received badge  Popular Question (source)
2015-03-31 02:18:13 -0500 asked a question dhcp dnsmasq.conf file for each network


I'm using Juno RDO, We have special needs for our dhcp servers in several tenant networks, the problem is, each network needs different options set up in the dhcp server. But we only have one file dnsmasq.conf where we can set the dhcp options for all dhcp servers/agents.

Is there a way to setup a dnsmasq.conf file per network?.

right now, what we do is configure several network nodes each one has a dnsmasq.conf file and is configured as the only dhcp agent for a network, but this way we can't have dhcp ha, and we also need several network nodes, not a good solution.


2015-03-31 02:04:30 -0500 received badge  Famous Question (source)
2015-03-31 01:59:55 -0500 received badge  Enthusiast
2015-03-30 08:37:54 -0500 received badge  Editor (source)
2015-03-30 08:37:13 -0500 asked a question modify hw_vif_model on deployed instances


I'm running juno, I need to modify the Property 'hw_vif_model' on and all-ready spawned image. I have modified the property on the glance image, but I don't want to delete and spawn the image again, is there a way to change this property on deployed instances ?.


2015-03-29 21:50:28 -0500 received badge  Notable Question (source)
2015-03-27 12:01:59 -0500 received badge  Popular Question (source)
2015-03-26 03:58:30 -0500 asked a question swift objects are not displayed with swift list


I have a very simple swift installation using Juno RDO, I have the proxy service on the controller and a couple of storage node, we just use the object storage, to store glance images and different type of backups.

It has working ok for some time now, but after a full reboot of the proxy and storage nodes, the problem we have is when we upload files to swift, after 50/60 of seconds they can't be listed anymore. The object still exists and I can do a GET an retrieve the object, the problem is the container and the objects are not listed anywhere:

[root@controller swift(keystone_fomvs)]# swift upload backup object.ring.gz ; echo "---" ; swift list backup; echo "---" ; sleep 60  ; swift list backup

Container 'backup' not found

Although the objects are not listed with the swift command, they can be retrieved without problems:

[root@controller swift(keystone_fomvs)]# swift stat
        Account: AUTH_4623b4b7b4a5451cac83b6a95fca898f
     Containers: 0
        Objects: 0
          Bytes: 0
X-Put-Timestamp: 1427358439.75538
    X-Timestamp: 1427358439.75538
     X-Trans-Id: tx1e7b5a2f634d4b7f96336-005513c2e7
   Content-Type: text/plain; charset=utf-8

[root@controller swift(keystone_fomvs)]# swift download backup object.ring.gz
object.ring.gz [auth 0.578s, headers 0.807s, total 0.808s, 0.005 MB/s]

No errors are reported in the swift logs, nor the rsync logs.

Here are the logs and the config files:

link:Link to pastebin

Could you please give me and idea where I can start looking to try and solve the problem?.


2015-01-14 08:55:57 -0500 asked a question cinder boot from lvm snapshot volume


I'm running IceHouse, and what I would like to do if possible is:

1.snapshot a cinder boot volume(lvm-iscsi backend). 2.boot the instance from the proper snapshot(not from a clone of the snapshot).

So right now, what I was able to do is:

  1. Create the snapshot of a boot volume:

    cinder snapshot-list +--------------------------------------+--------------------------------------+-----------+----------------+------+ | ID | Volume ID | Status | Display Name | Size | +--------------------------------------+--------------------------------------+-----------+----------------+------+ | 6071c186-2e6a-43cb-984e-51f3fca12ae8 | 46886de4-ce5f-4200-ac03-04132940cacf | available | fedora-os-snap | 8 | +--------------------------------------+--------------------------------------+-----------+----------------+------+

And then boot from a clone volume of the snapshot using:

nova boot --flavor m1.tindydisk --block-device-mapping vda=6071c186-2e6a-43cb-984e-51f3fca12ae8:snap::0 --key-name fomvs-key --nic net-id=a835bb18-d3fb-44d5-afd5-0c75887e4b21 --security-groups default fedora-test3

But what I need/would like, is to be able to boot from snapshot lvol directly:

[root@controller ~(keystone_fomvs)]# lvs LV VG Attr LSize Pool Origin Data% Move Log Cpy%Sync Convert _snapshot-6071c186-2e6a-43cb-984e-51f3fca12ae8 cinder-volumes swi-a-s--- 8.00g volume-46886de4-ce5f-4200-ac03-04132940cacf 0.00

let's say we have a classroom with 50 students, and we need to fire 50 instances, and we have very little space left in the cinder volumes(50 gigs). What I would do before(using lvm/virsh) is create 50 1gig lvol snapshots from a master, and then boot 50 Vms each from the 1 gig lvm snapshot.

Is it posible to do something like this with cinder/nova?, first create snapshots with different size from the original volume, and then tell nova to boot from those snapshots?.