Re: [Openstack] Urgent: How to recover a Controller?

fuel task list|grep -e 'pending|run' delete those tasks forcefully and then do delete-from-db --force again . add new node assign role controller.Deploy changes again as far as i know --delete-from-db --force doesnt need any deploy changes .It just disappeares from db and from "fuel node" list .

Re: [Openstack] RedHat OpenStack using Fuel- Fuel OpenStack release matrix

you can not have redhat as base os for environment deployed through fuel 9 or later.Earlier fuel used to support centos based deployment now its no longer available. On Fri, Apr 14, 2017 at 10:00 AM, pratik dave wrote: > Hi Team, > > I wanted to check what all OpenStack releases are possible to d

[Openstack] VRRP changes master to backup in juno

hi, I have 3 network nodes with min and max routers as 2. During testing i noticed that when a failed node comes online some of the master routers changes to backup . I checked configs and i dont see any difference in priority set for keepalived.conf but still active master changes to backup. Any

Re: [Openstack] High load average on network node for traffic of 100-200 kbps - Juno no DVR

looks like stopping dnsmasq bring down load average. Nothing specific to highlight in top and netstat ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.o

[Openstack] High load average on network node for traffic of 100-200 kbps - Juno no DVR

*Infrastructure* Running juno on ubuntu 14.04 with GRE network using 2 network nodes without DVR. There are around 200 dhcp namespaces created on each network nodes(dhcp_network_agents=2) which has 16 cpu and 32 gb. *Issue:* Network nodes are idle (top shows 99% idle CPU)but load average is *70-1

[Openstack] cinder-volume service HA

Hi is there any on going work to bring in native HA to cinder-volume service or is it already available in Kilo ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http:

Re: [Openstack] changing hostname by changing instance name

ot. If yes, then your cloud-init worked, just that metadata was > not updated by changing the instance name. If no, you need cloud-init built > in the image and it will poll the metadata after reboot. > > > > On 08/14/2015 10:45 AM, mad Engineer wrote: > > so is editing clou

Re: [Openstack] changing hostname by changing instance name

u will need to go into the instance if you do this after the > instance is booted. > > How to do so within the instance is OS specific. > On Aug 14, 2015 11:05 AM, "mad Engineer" wrote: > >> is there any way to change hostname by changing instance name. >> I

[Openstack] changing hostname by changing instance name

is there any way to change hostname by changing instance name. I tried changing instance name and then did a hard reboot but hostname is still the same.If this is not the right approach for changing hostname can some one tell me what should be done to change hostname of instances without logging to

[Openstack] BUG:Adding more tenants access to flavor with limited access causes already existing instance "size" in that tenant to "Not available" and this error pops up "Error: Unable to retrieve ins

Running Icehouse in Ubuntu 14.04 and Centos 6.6 (Both have this issue). Created a flavor "Flavor 1" with only "tenant A " having access to it and there are 20 instances created from that flavor. Now I have added 2 more tenants to access "Flavor 1" All the running instances in tenant A shows "Err

[Openstack] Multiple cinder lvm nodes

Read about cinder multi backend support,for using different backend types. I have 2 servers for running as lvm+iscsi back ends. Is it possible to use these 2 lvm backend nodes for same volume type and schedule block drives between these for that volume type. So that volumes will be distributed betw

[Openstack] Error: Unable to retrieve instance size information , even with correct flavor permission

Accidently edited flavor access permission while there were running instances from that flavor,now all the running instances show *Error: *Unable to retrieve instance size information. I can see all these instances with message in "admin " tenant with out any error, I can start new instance from

Re: [Openstack] iptables chain and instance id

7 | NULL | NULL | 0 | 25 | > 3 | 1bbb6888-b74f-4fc3-8c22-4c5231823567 | > > +-+++-++---+———+ > > The ID (25) corresponds to the chain name seen here: > > -A nova-compute-local -d 10.239.0.11/32 -j nova-compute-inst-25 >

Re: [Openstack] iptables chain and instance id

| >> >> +-+++-++---+--+ >> | 2013-07-03 14:40:47 | NULL | NULL | 0 | 25 | >> 3 | 1bbb6888-b74f-4fc3-8c22-4c5231823567 | >> >> +--

[Openstack] iptables chain and instance id

I am having issue troubleshooting iptables rules. How can i identify which chain belongs to which instance.. i can see nova-compute-inst-X but i am not able to relate X to nova list or to virsh list,Can some one please help in identifying proper iptables chains __

[Openstack] udev rules when starting instance from glance snapshot

Hello All, I have taken snapshot from running instances,when creating instance from that snapshot its failing to get new ip address,because of existing udev rules of eth0. How can this be fixed other than manually removing udev rules before creating snapshots. Thanks ___

[Openstack] Internal Server Error: /dashboard/project/volumes/ when deleting multiple volumes on centos6.6 icehouse

Hi, When deleting multiple volumes,dashboard shows "something went wrong" error,Apache error log has following error,no issue with deleting single volume.Request is not even reaching cinder-volume node. [Thu Feb 12 15:51:31 2015] [error] Internal Server Error: /dashboard/project/volumes/ [Thu F

Re: [Openstack] OverflowError: cannot convert float infinity to integer

/01/15 03:43, mad Engineer wrote: > > Hello All, > > Dashboard was working fine ,strangely after a failed > > instance resize and removing that instance,(instance-resize directory > > was till there) dashboard is returning http error code 500. > > > >

Re: [Openstack] higher MTU for all interfaces

; ------- > Date: Mon, 9 Feb 2015 15:01:57 +0530 > From: mad Engineer > To: Eren T?rkay > Cc: "openstack@lists.openstack.org" > Subject: Re: [Openstack] higher MTU for all interfaces > Message-ID: > > Content-Type: text/plain

Re: [Openstack] higher MTU for all interfaces

thanks,i manually changed br-tun br-int in its network config file and now its working across reboot. On Mon, Feb 9, 2015 at 12:42 PM, Eren Türkay wrote: > On 08-02-2015 21:56, mad Engineer wrote: >> Hello all is there any way we can change MTU of all relevant >> interfaces of

Re: [Openstack] higher MTU for all interfaces

, you'll still be left with > poor inter-instance communication and introduce some funky errors accessing > SSL sites from those instances to boot. > > -Erik > > On Feb 8, 2015 3:09 PM, "mad Engineer" wrote: >> >> Hello all is there any way we can change MTU

Re: [Openstack] higher MTU for all interfaces

, Remo Mattei wrote: > If you are using neutron you need to change it there > > ovs_neutron_plugin.ini agent veth_mtu 1500 for example. > > Remo >> On Feb 8, 2015, at 11:56, mad Engineer wrote: >> >> Hello all is there any way we can change MTU of all relevant &g

[Openstack] higher MTU for all interfaces

Hello all is there any way we can change MTU of all relevant interfaces of instances,if i set dhcp-option-force=26,1400 will it change MTU of all relevant interfaces like qbr qvo br-int etc etc is there any way to change MTU of all neutron interfaces. Thanks __

[Openstack] dhcp lease expired on instance with 100+days of uptime

hello all i have an instance with more than 100 days of uptime.it was mostly idle for months. For some reason instance was not responding to ping or any network packets. A Closer look inside vm shows that the dhcp ip received on interface has expired and a "dhclient" fixed everything (accessed th

[Openstack] Neutron network node with out using servers

Hello All, Is there any enterprise network device that can fully replace neutron network node? with all the features like tunneling,Functionality of separate network namespace,ie dhcp/router,Redundancy,Firewall rules for each tenant,VLAN isolation for each ip namespace. i am basical

Re: [Openstack] Live migration

are you doing live migration of instances on shared storage(looks like a non shared environment) do you have libvirtd listening on tcp port On Fri, Jan 16, 2015 at 10:02 PM, Paul Carlton wrote: > I've trying to do live migrations and am getting the following errors in the > libvirt_debug.log... >

[Openstack] multiple cinder backend with emc vnx and NFS backend

Hello All, i am working on integrating VNX with cinder,i have plan to add another NFS storage in the future,without removing VNX. Can i add another backend while first backend is running without causing problem to running volumes. I heard that multiple backend is supported, thanks f

Re: [Openstack] Auto sizing filesystem using cloud-init is not always working

Thanks gustavo,its working now :) On Thu, Jan 15, 2015 at 4:38 PM, gustavo panizzo (gfa) wrote: > > > On 01/15/2015 07:05 PM, mad Engineer wrote: >> >> Thanks gustavo, >> i have epel configured,through which i >> installed cloud-utils a

Re: [Openstack] Auto sizing filesystem using cloud-init is not always working

Thanks gustavo, i have epel configured,through which i installed cloud-utils and utils-growpart but not able to find "dracut-modules-growroot " is it part of some other rpm? On Thu, Jan 15, 2015 at 4:04 PM, gustavo panizzo (gfa) wrote: > > > On 01/15

Re: [Openstack] Auto sizing filesystem using cloud-init is not always working

at 3:24 PM, gustavo panizzo (gfa) wrote: > do you have installed dracut-module-growpart in the image? > have you rebuild the initramfs after installing it? > > On January 15, 2015 5:32:54 PM GMT+08:00, mad Engineer > wrote: >>hi i am using cloud-init on Centos6.6

[Openstack] Auto sizing filesystem using cloud-init is not always working

hi i am using cloud-init on Centos6.6 instances. Issue is instance's disk size is increasing but partition size not expanding according to flavor size i have growpart installed. growpart /dev/vda 1 shows following error NOCHANGE: partition 1 is 83883776.It can not be grown. But if i resize its w

[Openstack] OverflowError: cannot convert float infinity to integer

Hello All, Dashboard was working fine ,strangely after a failed instance resize and removing that instance,(instance-resize directory was till there) dashboard is returning http error code 500. error log shows: OverflowError: cannot convert float infinity to integer using icehouse on

Re: [Openstack] Allocated virtual RAM 512 MB even if no instance

This is based on reserved_host_memory_mb ,by default its 512. your instances can use only 31911-512 On Thu, Jan 8, 2015 at 5:18 PM, ppnaik wrote: > Hi All, > > On my compute node the nova-compute.log shows: > > 2015-01-08 17:13:19.995 88005 AUDIT nova.compute.resource_tracker [-] Total > physical

Re: [Openstack] ram_allocation_ratio per compute node or per AZ

gate with a ram ratio set > won't have a ram limit at all. > > You might also find the aggregate filter causes problems if you have a lot > of hosts, as it does a dB look up for each host per VM request. > > Phil > > On 20 Dec 2014 14:28, mad Engineer wrote: > Than

Re: [Openstack] ram_allocation_ratio per compute node or per AZ

s "Filter AggregateRamFilter returned 1 host(s)" " Filter RamFilter returned 0 hosts" and fails with error "instances:17 does not have 2048 MB usable ram, it only has 613.5 MB usable ram." Any idea? On Sat, Dec 20, 2014 at 3:09 PM, Antonio Messina wrote: > On

[Openstack] ram_allocation_ratio per compute node or per AZ

Hello All, I would like to know if its possible to set "ram_allocation_ratio" per compute node or at least per Availability zone. I tried setting that per compute nodes,but i get "no hosts found" message when allocation ratio reaches default 1.5 .Changing it on controller fixes this,b

Re: [Openstack] Automatically boot VM after server crash

there is a libvirt way to do this,using "virsh autostart" or creating symlinks manually,but i don't know what its impact in an openstack environment. On Tue, Dec 16, 2014 at 2:56 PM, abhishek jain wrote: > Hi Fiorenza > > Which parameter needs to be set true? I can check that at my end. > > On T

Re: [Openstack] High Latency to VMs

e8-ad > port 16: qvo714fab88-60 > port 17: qvob9ddde49-86 > port 18: qvo42ef9f3b-ac > port 19: qvof4ae7868-41 > port 20: qvoa4408a18-03 > port 22: qvo36c64d52-9b > > On 11 December 2014 at 06:17, mad Engineer wrote: >> >> sorry its 2.3.0 not 2.1.3 >> >> On Th

[Openstack] single 10G NIC and isolation in nova-network

hello All, I am trying to setup a test environment of icehouse,but the server has only one NIC. I want to achieve isolation of tenant traffic,management traffic,API and storage all go through this card. My doubts are : 1. If i configure to use trunk VLAN and cre

Re: [Openstack] High Latency to VMs

sorry its 2.3.0 not 2.1.3 On Thu, Dec 11, 2014 at 2:43 PM, mad Engineer wrote: > Not in openstack,i had performance issue, with OVS and bursty traffic > upgrading to later version improved the performance.A lot of > performance features have been added in 2.1.3. > > Do you ha

Re: [Openstack] High Latency to VMs

we are using version 2.0.2. > The process uses only about 0.3% on network node and compute node. > Did you have the same issue? > > On 10 December 2014 at 14:31, mad Engineer wrote: >> >> are you using openvswitch? which version? >> if yes,is it consuming a lot of CPU? &

Re: [Openstack] High Latency to VMs

are you using openvswitch? which version? if yes,is it consuming a lot of CPU? On Wed, Dec 10, 2014 at 7:45 PM, André Aranha wrote: > Well, here we are using de Icehouse with Ubuntu 14.04 LTS > > We found this thread in the community and we apply the changes in the > compute nodes (change VHOST_

[Openstack] QOS in neutron

hello All, i am using icehouse with neutron backend with open vswitch plugin. in my new flavor i have set qos as extra params : quota:vif_inbound_average": "2000" and i can see that in created vm xm file (virsh dumpxml) under "bandwidth" but how is it implemented in Neutron,i don't s

Re: [Openstack] significanse of reserved_host_memory_mb

ob/master/nova/compute/resource_tracker.py#L708 > [1] > https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L532 > > > > On 11/27/2014 01:55 PM, mad Engineer wrote: >> >> it reports "Free ram (MB): 425" >> but free -m has diff

Re: [Openstack] significanse of reserved_host_memory_mb

hould see in the logs the entire scheduler logic > and what resources it thinks your host has. > > On 27 Nov 2014 06:20, "mad Engineer" wrote: >> >> George, >> overcommit of RAM is 1 and that is working.However >> ins

Re: [Openstack] significanse of reserved_host_memory_mb

tances > using more virtual memory than the available physical memory on the host, > 700 MB in your case. > On 27 Nov 2014 05:36, "mad Engineer" wrote: > >> hi all i have set >> reserved_host_memory_mb in nova.conf of controller and compute and >> restarted n

[Openstack] significanse of reserved_host_memory_mb

hi all i have set reserved_host_memory_mb in nova.conf of controller and compute and restarted necessary services. i am expecting scheduler to not pickup host that has less than what is reserved_host_memory_mb in my example i put reserved_host_memory_mb = 1024 and free RAM in compute node is 700 Mb

[Openstack] using UUID instead of instance name

Hi, I am using icehouse with legacy network model ie nova-network and in this model instance name becomes DNS entry which was ok until tenants started creating instance with same name and started complaining about DNS issue.I am aware that in neutron it uses host-ip-address format and not inst

[Openstack] EMC VNX with QCOW2 disk images in cinder

Hi, Is there any way to use EMC VNX with out using vnx direct driver.I want to use qcow2 virtual disk and benefit from its thin provisioning and snapshot capability rather than spending more money on features which is already there in my hypervisor :-( .Is it possible? Experience with vmware a

Re: [Openstack] cinder and EMC VNX storage

thing provisioning ie vmdk thin provisioning over 3par thin provisioning. Thanks for your help On Sat, Nov 15, 2014 at 3:12 AM, yang, xing wrote: > Yes, VNX snapshot should work with thick LUN. > > > > Thanks, > > Xing > > > > > > *From:* mad Engineer [mailt

Re: [Openstack] cinder and EMC VNX storage

er. Maybe they are thick LUNs only and snapshots are snapview > snapshots instead of VNX snapshots. > > > > You need to have cinder-volume running, either on a separate node or on > the controller node in order to use a cinder driver. > > > > Thanks, > > Xing >

[Openstack] cinder and EMC VNX storage

Hi all, Currently i am using icehouse release with KVM compute nodes and a server with *cinder-volume* installed as my storage server,(it uses iscsi to export lvm disks to compute nodes.) Now i got a chance to TEST EMC storage ie EMC VNX for couple of days which is currently used by our vmw

[Openstack] floating ip list shows whole subnet for non-admin user

Hi, i am using nova-network and using this command trying to list available floating ip in my tenant: *nova-manage floating list* but this shows a whole subnet and not just floating ip available in my tenant. other commands like *nova list* is showing as expected. How can i restrict users t

[Openstack] external dns server for instances-"nova-network"

Hi, Is there any way to use external dns server for instances? All instances are currently getting instance name as hostname and with in openstack network its working and i manually add entries in corporate BIND to make it work for outside openstack. Is it possible to use an an external DNS se

[Openstack] dynamic VLAN in "nova-network"

hello All, I am using icehouse with *nova-network.* Is there a way to assign VLAN to all new tenants as soon tenant is created,probably by specifying a range of VLAN to be used. Currently i manually assign VLAN once tenant is created and its very difficulty as for each tenant a tick

[Openstack] neutron dhcp "scheduler" not working-icehouse

hi all, i have multiple network nodes (initially it was one and added one more node to meet demand) but what i noticed is "dhcp" namespaces are still created in old network node while l3 scheduling is working as i can see router getting created on new nodes. *neutron ext-list -c name -c ali

[Openstack] how does metadata server get information about "user data"

Hi, I am trying to understand more about metadata server. How will metadata server running on network node gets information about what we write in Post Creation> customisation script . How will it download from dashboard to metadata server and where does it store that file. Can some one explai

[Openstack] used floating_ip counter not decreasing even after removing instance

hi in my setup i have created 10 instances each with floating_ip and deleted all instances,*with out disassociating floating ip.* Now when i try to associate floating ip with new instance,it shows "unable to associate floating ip" and in dashboard used floating_ip is 10/10 even with only 5 insta

[Openstack] nova-network and loadbalancers

Hi, Is there any way to make HA proxy load balancer work with nova-network,worried about fail over ip conflicting with IP/MAC stealing rule. trying to configure 2 HA proxy with keepalived and a fail over IP. OR Is there a load balancer that can actually work with nova-network Thanks, ___

[Openstack] legacy nova-network and auto scaling

Hi, I have working ice house release with "nova-network" as the networking service. My question is,is there any dependency on auto scaling and neutron ie do i need neutron for auto scaling. Thanks ___ Mailing list: http://lists.openstack.org/cgi-bin/

Re: [Openstack] even with 60G qcow2 image " FlavorDiskTooSmall: Flavor's disk is too small for requested image. " for m1.small

Thanks Don, it works, i was thinking that image should be large enough to fit in all flavor sizes ;-) On Sun, Aug 10, 2014 at 8:39 PM, Don Waterloo wrote: > On 10 August 2014 09:55, mad Engineer wrote: > > > > Hi, > > i am using Centos6.5 with Ic

[Openstack] even with 60G qcow2 image " FlavorDiskTooSmall: Flavor's disk is too small for requested image. " for m1.small

Hi, i am using Centos6.5 with Icehouse release,on KVM Hypervisor, Trying to launch Centos instance from a newly uploaded qcow2 image. Created centos6.5 image in qcow2 format with 60G virtual size and actual size of 1.7G image: Centos6.5_x64.qcow2 file format: qcow2 virtual size: 60G (644245

Re: [Openstack] nova-network,VRRP and NAT

Thanks Xav, i am using nova-network and not neutron.Looks like this can not work with nova-network Thanks On Thu, Aug 7, 2014 at 3:23 PM, Xav Paice wrote: > On 07/08/14 21:42, mad Engineer wrote: > > but concerned whether nova security policies allow VRRP to w

[Openstack] nova-network,VRRP and NAT

Hi, I am using nova-network on Havana in multi-node setup with almost 20 instances that are web servers. i am planning to use HA proxy cluster with keepalived for fail over. but concerned whether nova security policies allow VRRP to work as it requires multiple IP on same MAC? Is cle