Re: [Openstack-operators] [openstack-dev] [nova] How to expose the compute node local disks to instances

Hi Belmiro, In Cinder there is the "raw disk device" driver that has been used by some for Hadoop and similar applications. That may be something to look at, but with a big warning. That being that it is deprecated in the Ocata release and will be removed in Pike. The reason it is going to be re

[Openstack-operators] [openstack-dev] [nova] How to expose the compute node local disks to instances

Hi, we have a set of disk servers (JBOD) that we would like to integrate into our cloud to run applications like Hadoop and Spark. Using file disks for storage and a huge "/var/lib/nova" is not an option for these use cases so we would like to expose the local drives directly to the VMs as ephe

Re: [Openstack-operators] Any serious stability and performance issues on RBD as ephemeral storage ?

I'd say using Ceph for ephemeral disks is the most common deployment pattern, at DataCentred we've been using it for years. It's rock solid and has been for several releases. There were some edge case issues around resizing and snapshotting, but I think that's all been fixed in the last couple of i

Re: [Openstack-operators] Any serious stability and performance issues on RBD as ephemeral storage ?

We've been using it and recommending it for years. It solves many many problems with a running cloud and there have been very few issues. Pay close attention when upgrading versions of CEPH and do things in the right order and you will be fine! On Wed, Dec 7, 2016 at 7:51 AM, Vahric Muhtaryan wro

Re: [Openstack-operators] Any serious stability and performance issues on RBD as ephemeral storage ?

On 12/7/2016 9:25 AM, Jonathan Proulx wrote: We've been using Ceph as ephemeral backend (and glance store and cinder backend) for > 2 years (maybe 3 ) and have been very happy. cinder has been rock solid on RBD side. Early on when we had 6 osd servers we lost one in production to a memory error

Re: [Openstack-operators] [cinder] os-vol-host-attr:host value invalid after upgrade

Ignore this, it's not the upgrade of course, but a config change. It looks like this is something un-accounted for in the puppet module migrations that have taken place as well, and so easily fixed :) Cheers, Just On 8 December 2016 at 14:01, Justin Cattle wrote: > Hi, > > > After migrating

Re: [Openstack-operators] Mitaka to Newton networking issues

Hi Grant, Sorry to hear that. Can you do me a favor and file a bug against neutron with your env so we can try to reproduce it? Include everything like if you have l2pop enabled, which agent you are using, and which L3 settings you are using (L3HA/DVR/etc). Thanks, Kevin Benton On Thu, Dec 8, 20

[Openstack-operators] [cinder] os-vol-host-attr:host value invalid after upgrade

Hi, After migrating from Juno to Mitaka [ via Kilo and Liberty in between ], new cinder-volume services are created with pool specific names, instead of host specific names. e.g. Old service: node123@rbd-volumes#rbd-volumes New service: rbd:poolzone1@rbd-volumes#rbd-volumes The old se

Re: [Openstack-operators] [scientific] Computing Insight UK in Manchester - Scientific OpenStack breakout session?

I have arranged a Scientific OpenStack breakout session for this conference - 2pm on Thursday 15th. If you’re coming to CIUK [https://eventbooking.stfc.ac.uk/news-events/ciuk-2016 ], please do come along - and also let me know if you would

[Openstack-operators] Inventory contains out of date data

Hi, This is partly related to a previous submission "Mitaka to Newton networking issues" by a colleague of mine, Grant Morley. We had to rebuild our ansible server due to issues experienced when reverting from the Newton to Mitaka versions, which left ansible in an unworkable state. Our osa (/

Re: [Openstack-operators] Mitaka to Newton networking issues

Hi Kevin, Sorry for the late reply, We have tried doing that and we were still seeing the same issues. I don't think the bug was quite the same as what we were seeing. Unfortunately we have had to roll back to Mitaka as we had a tight deadline and not being able to create networks / have HA

Re: [Openstack-operators] [User-committee] [publicClouds-wg] Reminder: Public Cloud WG meeting Wednesday 1400UTC

As always, great input Flanders - and thank you for adding to our etherpad. Our first meeting was well attended, with 10+ public cloud operators represented. We've made progress with some of the initial housekeeping work, and will be looking at defining some of the purpose and scope next meeting. I