Re: [Openstack] Struggling to get the s3 api interface to work with swift.

[John van Ommen]

What release are you using?

In 2016 I tried to get this working for a client of mine at HPE, and found
that it wouldn't work without a fair bit of hacking. Basically the software
hadn't been updated in about a year, and the newest release was
incompatible with the version of OpenStack that we were selling.

On Thu, May 24, 2018 at 11:57 PM, Shyam Prasad N 
wrote:

> Thanks. I'll try this.
> But what values do I use in place of ak and sk? I want to use some command
> to get those values, right?
>
> On Fri, May 25, 2018 at 9:52 AM, Yuxin Wang 
> wrote:
>
>> I created ec2 credentials using command `openstack credential create`.
>>
>> i.e.
>>
>> openstack credential create --type ec2 --project proj user '{"access":
>> "ak", "secret": "sk”}'
>>
>>
>> It seems the two credentials are not the same thing.
>>
>> Ref:
>>
>> https://www.ibm.com/support/knowledgecenter/en/STXKQY_4.1.1/
>> com.ibm.spectrum.scale.v4r11.adv.doc/bl1adv_ConfigureOpenst
>> ackEC2credentials.htm
>>
>> 在 2018年5月25日，10:32，Shyam Prasad N  写道：
>>
>> Yes, I did.
>> I don't think this is s3curl related issue, because I tried with python
>> AWS SDK, and got the same error.
>>
>> On Fri, May 25, 2018, 07:42 Yuxin Wang 
>> wrote:
>>
>>> Did you add 127.0.0.1 to the endpoint list in s3curl.pl?
>>>
>>> i.e.
>>>
>>> my @endpoints = (‘127.0.0.1’);
>>>
>>> 在 2018年5月24日，19:48，Shyam Prasad N  写道：
>>>
>>> Hi,
>>>
>>> I've been trying to get swift3 to work for several days now. But I
>>> haven't managed to get it running.
>>> Both with tempauth and keystoneauth, I'm getting the same error:
>>>
>>> eightkpc@objectstore1:~/s3curl$ ./s3curl.pl
>>> 
>>> --id=testerks -- http://127.0.0.1:8080/
>>> 
>>> 
>>> SignatureDoesNotMatchThe request signature
>>> we calculated does not match the signature you provided. Check your key and
>>> signing method.txa691e7ca97a44d56bc4c2-005b06a2
>>> 92
>>>
>>> May 24 11:31:30 localhost proxy-server: 127.0.0.1 127.0.0.1
>>> 24/May/2018/11/31/30 GET / HTTP/1.0 403 - curl/7.58.0 - - 277 -
>>> txa691e7ca97a44d56bc4c2-005b06a292 - 0.0200 - - 1527161490.543112040
>>> 1527161490.563107014 -
>>> May 24 11:31:30 localhost proxy-server: STDERR: 127.0.0.1 - -
>>> [24/May/2018 11:31:30] "GET / HTTP/1.1" 403 621 0.021979 (txn:
>>> txa691e7ca97a44d56bc4c2-005b06a292)
>>>
>>> eightkpc@objectstore1:~$ cat .s3curl
>>> %awsSecretAccessKeys = (
>>> tester => {
>>> id => 'test:tester',
>>> key => 'testing',
>>> },
>>> testerks => {
>>> id => 'e6289a1b5692461388d0597a4873d054',
>>> key => '88bb706887094696b082f008ba133ad7',
>>> },
>>> );
>>>
>>> eightkpc@objectstore1:~$ openstack ec2 credentials show
>>> e6289a1b5692461388d0597a4873d054
>>> ++--
>>> 
>>> --+
>>> | Field  | Value
>>>
>>>  |
>>> ++--
>>> 
>>> --+
>>> | access | e6289a1b5692461388d0597a4873d0
>>> 54
>>>  |
>>> | links  | {u'self': u'http://controller:5000/v3/us
>>> ers/d7df7b56343b4ea988869fc30efeda09/credentials/OS-EC2/e628
>>> 9a1b5692461388d0597a4873d054'} |
>>> | project_id | dc86f7d8787b46158268bd77098b65
>>> 78
>>>  |
>>> | secret | 88bb706887094696b082f008ba133a
>>> d7
>>>  |
>>> | trust_id   | None
>>>
>>>  |
>>> | user_id| d7df7b56343b4ea988869fc30efeda
>>> 09
>>>  |
>>> ++--
>>> 
>>> --+
>>>
>>> Can someone please let me know what is going on?
>>>
>>> Regards,
>>> Shyam
>>> ___
>>> Mailing list: https://eur03.safelinks.protec
>>> tion.outlook.com/?url=http%3A%2F%2Flists.openstack.org%
>>> 2Fcgi-bin%2Fmailman%2Flistinfo%2Fopenstack=02%7C01%7C%7
>>> C39742b8c6bf847ee381508d5c16d1b21%7C84df9e7fe9f640afb435
>>> %7C1%7C0%7C636627596701206160=KI%2F2T2FhVQJTeX
>>> 1KbIObDZVDiUA3SbTq6Pplo1bc7ak%3D=0
>>> Post to : 

Re: [Openstack-operators] 4K block size

[John van Ommen]

The benchmarks DO appear to show a modest improvement.

In the bugzilla post you linked, Paolo said that

"With 4k logical sector size in the host, you must have a 4k logical
sector size in the guest too."

So it appears you'd need to use a disk that physically supports it, along
with an OS that supports it also.

RHEL7 supports it, here's some detail:

https://access.redhat.com/solutions/56494



On Mon, Apr 23, 2018, 10:47 AM Tim Bell  wrote:

>
>
> Has anyone experience of working with local disks or volumes with
> physical/logical block sizes of 4K rather than 512?
>
>
>
> There seems to be KVM support for this (
> http://fibrevillage.com/sysadmin/216-how-to-make-qemu-kvm-accept-4k-sector-sized-disks)
> but I could not see how to get the appropriate flavors/volumes in an
> OpenStack environment?
>
>
>
> Is there any performance improvement from moving to 4K rather than 512
> byte sectors?
>
>
>
> Tim
>
>
> ___
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

Re: [Openstack] Solaris on openstack

[John van Ommen]

You've seen this, right?

https://dcvan24.wordpress.com/2013/07/11/create-a-solaris-10-x86-image-for-openstack/

"*2. Install Solaris*

First of all, create a QCOW2 image. There are several installation options
for Solaris 10, but believe me you want a developer’s installation. You
want bash, you want SSH, say yes… A developer’s installation needs about
6GB space, so it’s safe to create a image larger than 6GB
1
qemu-img create -f qcow2 solaris.qcow2 7G

Then you need a libvirt.xml file to start the VM. libvirt.xml is a
configuration file for a VM. You can define everything about the VM in this
file, including the number of cores, memory size, disk, network, etc. The
following listing is a template of libvirt.xml used for Solaris
installation."

On Aug 14, 2017 7:01 AM, "John van Ommen" <john.vanom...@gmail.com> wrote:

> Take a look at this:
>  https://pve.proxmox.com/wiki/Solaris_Guests
>
> And this:
> https://www.linux-kvm.org/page/Guest_Support_Status#UNIX_Family:_Solaris.
> 2FOpenSolaris
>
> In a nutshell, it looks like all versions of Solaris 10 have various
> issues running under kvm but there are workarounds.
>
> Some releases require modification of the grub parameters to get Solaris
> to boot under kvm.
>
>
>
>
> On Aug 14, 2017 6:45 AM, "Shyam Biradar" <shyambiradarsgg...@gmail.com>
> wrote:
>
> Hello John,
>
> I am using KVM hypervisor with that I want to launch solaris 10 OS as
> guest vm using its ISO image.
> I am able to start installation process but at the final step its failing
> to find installation media through cd/dvd device.
> Attached screenshot.[solaris.JPG]
>
> Error after selecting cd/dvd option: No disks found
>
>
> I tried some other ways like launching solaris on esx and converting its
> vmdk to qcow2, with this approach qcow2 boot process is getting stuck at
> very initial stage(at boot options step).
> Attached screenshot{solaris-qcow2.JPG]
>
>
> Thanks & Regards,
> Shyam Biradar,
> Email: shyambiradarsgg...@gmail.com,
> Contact: +91 8600266938 <+91%2086002%2066938>.
>
> On Mon, Aug 14, 2017 at 3:53 PM, John van Ommen <john.vanom...@gmail.com>
> wrote:
>
>> What is the error that you are receiving?
>>
>> On Mon, Aug 14, 2017 at 1:22 AM, Shyam Biradar <
>> shyambiradarsgg...@gmail.com> wrote:
>>
>>> Hi,
>>>
>>> I am trying to install solaris on OpenStack. I am facing many issues
>>> during os installation process.
>>>
>>> Does OpenStack support solaris in any release/distribution?
>>>
>>>
>>> Thanks & Regards,
>>> Shyam Biradar,
>>> Email: shyambiradarsgg...@gmail.com,
>>> Contact: +91 8600266938 <+91%2086002%2066938>.
>>>
>>> ___
>>> Mailing list: http://lists.openstack.org/cgi
>>> -bin/mailman/listinfo/openstack
>>> Post to : openstack@lists.openstack.org
>>> Unsubscribe : http://lists.openstack.org/cgi
>>> -bin/mailman/listinfo/openstack
>>>
>>>
>>
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Solaris on openstack

[John van Ommen]

What is the error that you are receiving?

On Mon, Aug 14, 2017 at 1:22 AM, Shyam Biradar  wrote:

> Hi,
>
> I am trying to install solaris on OpenStack. I am facing many issues
> during os installation process.
>
> Does OpenStack support solaris in any release/distribution?
>
>
> Thanks & Regards,
> Shyam Biradar,
> Email: shyambiradarsgg...@gmail.com,
> Contact: +91 8600266938 <+91%2086002%2066938>.
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack-operators] Tricircle

[John van Ommen]

My team at HPE did a few projects like that:

https://saas.hpe.com/en-us/software/cloud-service-automation

I believe BMC also offers products in that space, but you'd have to contact
them to find out what they're capable of.

For the past year, most customers I've talked to have been leaning towards
container solutions. IE, instead of buying a product that can provision
resources in two disparate clouds, they're deploying containers and using a
container orchestration product to bridge the environments. (Figuratively,
not literally.)

Hybrid cloud is not a trivial endeavor.

John



On Tue, Jul 11, 2017 at 11:10 AM, Curtis  wrote:

> On Tue, Jul 11, 2017 at 10:47 AM, Ignazio Cassano
>  wrote:
> > I would like bothnetworking is important but an heat multi pod
> > Orchestrator could be fantastic.
>
> So like a higher level system that can manage multiple clouds via
> their heat API?
>
> I'm only familiar with some work in the NFV area around MANO
> (management and orchestration). There are several systems that can
> manage multiple clouds, some using heat and others using the standard
> APIs. One OpenStack related example would be the Tacker system.
>
> I think a higher level heat system that could manage other heat
> systems would be interesting. I see some mention of heat multicloud
> but I'm not sure where that ended up. I should look into that...
>
> Thanks,
> Curtis.
>
>
> > Regards
> > Ignazio
> >
> > Il 11/Lug/2017 06:20 PM, "Curtis"  ha scritto:
> >>
> >> On Tue, Jul 11, 2017 at 9:54 AM, Ignazio Cassano
> >>  wrote:
> >> > Hi openstackers,
> >> > anyone is using tricircle in production environment?
> >> > Any alternative for e multi pod openstack like tricircle?
> >>
> >> What is it you want to do?
> >>
> >> AFAIK tricircle has pivoted recently to accomplish networking across
> >> multi-region openstack deployments.
> >>
> >> Are you mostly looking for networking across clouds or are you looking
> >> to tie a bunch of clouds together with some higher level abstraction?
> >> Or both. :)
> >>
> >> Thanks,
> >> Curtis.
> >>
> >> > Regards
> >> > Ignazio
> >> >
> >> >
> >> > ___
> >> > OpenStack-operators mailing list
> >> > OpenStack-operators@lists.openstack.org
> >> > http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack-operators
> >> >
> >>
> >>
> >>
> >> --
> >> Blog: serverascode.com
>
>
>
> --
> Blog: serverascode.com
>
> ___
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

Re: [Openstack] Production openstack on 5 node

[John van Ommen]

I couldn't agree more. OpenStack deployment is not a trivial task. Do not
reinvent the wheel.

John

On Jun 21, 2017 12:04 PM, "Erik McCormick" <emccorm...@cirrusseven.com>
wrote:

> If you are a Puppet shop you should check out the Puppet community
> modules. You will get familiar with all the inner goo as you'll need your
> own composition layer (in my case a series of hiera files). There's no
> reason to reinvent the wheel unless you really want to.
>
> -Erik
>
> On Jun 21, 2017 2:19 PM, "Satish Patel" <satish@gmail.com> wrote:
>
>> Thank you all of you for your opinion, As mike suggested i am also
>> planning to brew home grown puppet module to understand each and every
>> component and their role instead of grabbing third party tool, I am
>> sure OOO is best for 100 deploying 100 compute node but in my setup we
>> have only 5 servers and its not worth it to manage undercloud server.
>>
>> We are puppet shop so it would be easy to write own code and go from
>> there.
>>
>> On Wed, Jun 21, 2017 at 1:25 AM, Remo Mattei <r...@italy1.com> wrote:
>> > I did a deployment with cs9 hp it was pretty bad. I hope the new one
>> does
>> > better.
>> >
>> > Nevertheless I do not see many using hp out there. Maybe different
>> regions
>> > like emea do better with that.
>> >
>> > Inviato da iPhone
>> >
>> > Il giorno 20 giu 2017, alle ore 21:10, John van Ommen
>> > <john.vanom...@gmail.com> ha scritto:
>> >
>> > At HPE we originally used TripleO but switched to a 'flat' model.
>> >
>> > I personally didn't see any advantage to Triple O. In theory, it should
>> be
>> > easier to manage and upgrade. In the real world, Helion 3.0 and 4.0 are
>> > superior in every respect.
>> >
>> > John
>> >
>> > On Jun 20, 2017 9:02 PM, "Remo Mattei" <r...@italy1.com> wrote:
>> >>
>> >> I worked for Red Hat and they really want to get ooo going because the
>> >> installation tools did never work as everyone was hoping. Before Red
>> Hat I
>> >> was at Mirantis and the fuel installation was nice now dead. I know
>> ooo will
>> >> go into containers next couple of release but kolla–Ansible is one of
>> the
>> >> emerging solutions now to get it out fast.
>> >>
>> >> I am doing a project now where I am working on deploying ooo just
>> finished
>> >> the doc for Ocata undercloud.
>> >>
>> >> Just my two cents to concord with Mike’s statement.
>> >>
>> >> Remo
>> >>
>> >> Inviato da iPhone
>> >>
>> >> Il giorno 20 giu 2017, alle ore 20:51, Mike Smith <
>> mism...@overstock.com>
>> >> ha scritto:
>> >>
>> >> There are definitely 1,001 opinions on what is “best”.  We use RDO at
>> >> Overstock and we use home-grown puppet modules because we do our own
>> puppet
>> >> modules for everything else we do here.  We based everything around the
>> >> official Openstack install documents and we do it because we want to
>> *fully*
>> >> understand everything we can instead of treating it like a black box
>> that
>> >> knows how to do the magic.
>> >>
>> >> However, there are lots of options out there - ansible, kolla, puppet
>> plus
>> >> vendor-specific options too like those provided by Mirantis.  If there
>> are
>> >> config management tools (ansible, puppet, etc) that you already use,
>> you may
>> >> want to check out the Openstack options for those.  You are correct
>> that
>> >> that packstack is more of a ‘all-in-one-server’ installer for a quick
>> POC.
>> >> It can to more, but I think RDO recommends “Triple-O” (which stands for
>> >> Openstack-on-Openstack) for production RDO deploys.  Since they are
>> >> affiliated with RedHat, they would also lean heavily towards the
>> Ansible
>> >> option as well.
>> >>
>> >> Good luck!
>> >>
>> >>
>> >> Mike Smith
>> >> Overstock Cloud Team
>> >>
>> >>
>> >>
>> >> On Jun 20, 2017, at 5:49 PM, Erik McCormick <
>> emccorm...@cirrusseven.com>
>> >> wrote:
>> >>
>> >> This is a religious discussion for most, but I would suggest Kolla. It
>> >> takes out a lot of the guess work, h

Re: [Openstack] Production openstack on 5 node

[John van Ommen]

At HPE we originally used TripleO but switched to a 'flat' model.

I personally didn't see any advantage to Triple O. In theory, it should be
easier to manage and upgrade. In the real world, Helion 3.0 and 4.0 are
superior in every respect.

John

On Jun 20, 2017 9:02 PM, "Remo Mattei"  wrote:

> I worked for Red Hat and they really want to get ooo going because the
> installation tools did never work as everyone was hoping. Before Red Hat I
> was at Mirantis and the fuel installation was nice now dead. I know ooo
> will go into containers next couple of release but kolla–Ansible is one of
> the emerging solutions now to get it out fast.
>
> I am doing a project now where I am working on deploying ooo just finished
> the doc for Ocata undercloud.
>
> Just my two cents to concord with Mike’s statement.
>
> Remo
>
> Inviato da iPhone
>
> Il giorno 20 giu 2017, alle ore 20:51, Mike Smith 
> ha scritto:
>
> There are definitely 1,001 opinions on what is “best”.  We use RDO at
> Overstock and we use home-grown puppet modules because we do our own puppet
> modules for everything else we do here.  We based everything around the
> official Openstack install documents and we do it because we want to
> *fully* understand everything we can instead of treating it like a black
> box that knows how to do the magic.
>
> However, there are lots of options out there - ansible, kolla, puppet plus
> vendor-specific options too like those provided by Mirantis.  If there are
> config management tools (ansible, puppet, etc) that you already use, you
> may want to check out the Openstack options for those.  You are correct
> that that packstack is more of a ‘all-in-one-server’ installer for a quick
> POC.  It can to more, but I think RDO recommends “Triple-O” (which stands
> for Openstack-on-Openstack) for production RDO deploys.  Since they are
> affiliated with RedHat, they would also lean heavily towards the Ansible
> option as well.
>
> Good luck!
>
>
> Mike Smith
> Overstock Cloud Team
>
>
>
> On Jun 20, 2017, at 5:49 PM, Erik McCormick 
> wrote:
>
> This is a religious discussion for most, but I would suggest Kolla. It
> takes out a lot of the guess work, has a good upgrade mechanism, and is
> well supported by the community via mailing list and ORC. Take a look.
>
> -Erik
>
>
> On Jun 20, 2017 7:15 PM, "Satish Patel"  wrote:
>
>> We are deploying 5 node openstack cloud for internal use and wondering
>> what method we should use, initial test was on RDO packstack but i
>> heard packstack isn't good for production, some people on google
>> suggesting using triplo. I found its little complicated because you
>> need one more extra server run undercloud openstack to deploy
>> overcloud openstack, should i really use triplo or is there any other
>> and easy method which can allow us to upgrade in future also.
>>
>> ___
>> Mailing list: http://lists.openstack.org/cgi
>> -bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe : http://lists.openstack.org/cgi
>> -bin/mailman/listinfo/openstack
>>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
>
>
> !DSPAM:1,5949ed75173232896912281!
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
>
> !DSPAM:1,5949ed75173232896912281!
>
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] recommended way to install openstack today

[John van Ommen]

Hewlett Packard's "Helion Open Stack" has improved a great deal since
the initial release. They offer quite a few deployment options now,
and the reliability is quite good:

https://docs.hpcloud.com/hos-4.x/index.html#helion/installation/installation_overview.html

On Sun, Jan 15, 2017 at 2:38 PM, Manuel Sopena Ballesteros
 wrote:
> Hi,
>
>
>
> We have been testing openstack for a few days in a all-in-one flavor. Now we
> would like to have an openstack HA  production environment. I see there are
> multiple ways to do this… TripleO, fuel, kolla… my question is what would be
> your preferred way to deploy openstack today?
>
>
>
> Thank you very much
>
>
>
> Manuel Sopena Ballesteros | Big data Engineer
> Garvan Institute of Medical Research
> The Kinghorn Cancer Centre, 370 Victoria Street, Darlinghurst, NSW 2010
> T: + 61 (0)2 9355 5760 | F: +61 (0)2 9295 8507 | E: manuel...@garvan.org.au
>
>
>
> NOTICE
> Please consider the environment before printing this email. This message and
> any attachments are intended for the addressee named and may contain legally
> privileged/confidential/copyright information. If you are not the intended
> recipient, you should not read, use, disclose, copy or distribute this
> communication. If you have received this message in error please notify us
> at once by return email and then delete both messages. We accept no
> liability for the distribution of viruses or similar in electronic
> communications. This notice should not be removed.
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [OpenStack] [Swift] How to install Swift3 middleware?

[John van Ommen]

To me, the errors you're receiving are indicative of the swift3
package not installing correctly.

When you ran that first step (setup.py), did you get errors?

John

On Wed, Oct 12, 2016 at 10:03 PM, Alexandr Porunov
 wrote:
> Hello,
>
> It still doesn't work... The same error is shown. Maybe I missed something?
> Maybe before the installation I had to done something or something else. I
> tried different things to run it. But when we do have swift3 or/and s3token
> in pipeline it doesn't work.
>
> Sincerely,
> Alexandr
>
> On Thu, Oct 13, 2016 at 12:17 AM, Mark Kirkwood
>  wrote:
>>
>>
>>
>> On 13/10/16 06:26, Alexandr Porunov wrote:
>>>
>>> Hello,
>>>
>>> I want to use S3 api for Swift.
>>>
>>> I have installed swift3 as here:
>>>
>>> git clone https://github.com/openstack/swift3
>>> cd swift3
>>> python setup.py install
>>>
>>> Then I have configured my proxy-server.conf as here:
>>>
>>> [pipeline:main]
>>> pipeline = catch_errors gatekeeper healthcheck proxy-logging cache
>>> container_sync bulk ratelimit swift3 s3token authtoken keystoneauth
>>> container-quotas account-quotas slo dlo versioned_writes proxy-logging
>>> proxy-server
>>>
>>> [filter:swift3]
>>> use = egg:swift3#swift3
>>>
>>> [filter:s3token]
>>> use = egg:swift3#s3token
>>> auth_uri = http://keystone:35357/
>>> reseller_prefix = AUTH_
>>>
>>> Then I tried to restart the proxy:
>>> systemctl restart openstack-swift-proxy
>>>
>>> And it failes.
>>>
>>>
>>
>> Yeah - for Mitaka (or thereabouts) onwards that s3token config needs to
>> change:
>>
>> [filter:s3token]
>> paste.filter_factory = keystoneclient.middleware.s3_token:filter_factory
>> auth_uri = http://keystone:35357/
>> reseller_prefix = AUTH_
>>
>> Alternatively the older port/protocol/host variables still work instead of
>> auth_uri:
>>
>> [filter:s3token]
>> paste.filter_factory = keystoneclient.middleware.s3_token:filter_factory
>> auth_port = 35357
>> auth_host = keystone
>> auth_protocol = http
>>
>> The swift3 docs need updating, as it is not obvious what has gone wrong
>> when you hit this.
>>
>> regards
>>
>> Mark
>>
>>
>>
>>
>>
>>
>> ___
>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>
>
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [OpenStack] [Swift] How to install Swift3 middleware?

[John van Ommen]

Are you installing OpenStack from trunk?

I've installed Swift3 on Helion Open Stack 2.0, and the published
instructions didn't work for me at all.

On Wed, Oct 12, 2016 at 10:26 AM, Alexandr Porunov
 wrote:
> Hello,
>
> I want to use S3 api for Swift.
>
> I have installed swift3 as here:
>
> git clone https://github.com/openstack/swift3
> cd swift3
> python setup.py install
>
> Then I have configured my proxy-server.conf as here:
>
> [pipeline:main]
> pipeline = catch_errors gatekeeper healthcheck proxy-logging cache
> container_sync bulk ratelimit swift3 s3token authtoken keystoneauth
> container-quotas account-quotas slo dlo versioned_writes proxy-logging
> proxy-server
>
> [filter:swift3]
> use = egg:swift3#swift3
>
> [filter:s3token]
> use = egg:swift3#s3token
> auth_uri = http://keystone:35357/
> reseller_prefix = AUTH_
>
> Then I tried to restart the proxy:
> systemctl restart openstack-swift-proxy
>
> And it failes.
>
> systemctl status openstack-swift-proxy -l
> shows next:
>
> ● openstack-swift-proxy.service - OpenStack Object Storage (swift) - Proxy
> Server
>Loaded: loaded (/usr/lib/systemd/system/openstack-swift-proxy.service;
> enabled; vendor preset: disabled)
>Active: failed (Result: exit-code) since Wed 2016-10-12 23:17:06 EEST;
> 3min 0s ago
>   Process: 2213 ExecStart=/usr/bin/swift-proxy-server
> /etc/swift/proxy-server.conf (code=exited, status=1/FAILURE)
>  Main PID: 2213 (code=exited, status=1/FAILURE)
>
> Oct 12 23:17:06 proxy2 systemd[1]: Started OpenStack Object Storage (swift)
> - Proxy Server.
> Oct 12 23:17:06 proxy2 systemd[1]: Starting OpenStack Object Storage (swift)
> - Proxy Server...
> Oct 12 23:17:06 proxy2 liberasurecode[2213]: liberasurecode_backend_open:
> dynamic linking error libJerasure.so.2: cannot open shared object file: No
> such file or directory
> Oct 12 23:17:06 proxy2 liberasurecode[2213]: liberasurecode_backend_open:
> dynamic linking error libJerasure.so.2: cannot open shared object file: No
> such file or directory
> Oct 12 23:17:06 proxy2 liberasurecode[2213]: liberasurecode_backend_open:
> dynamic linking error libisal.so.2: cannot open shared object file: No such
> file or directory
> Oct 12 23:17:06 proxy2 liberasurecode[2213]: liberasurecode_backend_open:
> dynamic linking error libshss.so.1: cannot open shared object file: No such
> file or directory
> Oct 12 23:17:06 proxy2 systemd[1]: openstack-swift-proxy.service: main
> process exited, code=exited, status=1/FAILURE
> Oct 12 23:17:06 proxy2 systemd[1]: Unit openstack-swift-proxy.service
> entered failed state.
> Oct 12 23:17:06 proxy2 systemd[1]: openstack-swift-proxy.service failed.
>
>
> If I remove "swift3 s3token" form pipeline then it works.. Why my proxy
> crashes with swift3?
>
> I have an OpenStack Swift Mitaka
>
> Sincerely,
> Alexandr
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack-operators] SDN for hybridcloud, does it *really* exist?

[John van Ommen]

In regards to your last comment, that "it would be great for tenants
to be able to setup connections into AWS", HP CSA comes close to doing
that:

http://www8.hp.com/us/en/software-solutions/cloud-service-automation/

It's not *exactly* what you're looking for, you wouldn't be using the
OpenStack API. But it's as close as you can get right now.

IMHO, HP's OpenStack and it's integration with HP's automation tools
has made giant strides in the last year.

On Mon, Oct 3, 2016 at 4:29 PM, Curtis  wrote:
> On Mon, Oct 3, 2016 at 11:52 AM, Jonathan Proulx  wrote:
>>
>> So my sense from responses so far:
>>
>> No one is doing unified SDN solutions across clouds and no one really
>> wants to.
>
> I do want to (but am not doing). When I worked at a public cloud based
> on openstack we certainly wanted this kind of functionality, mostly
> between openstack regions. Likely so would telecoms, hence the
> tricircle link that was sent. But I left before we really got into it
> so I'm not sure what's happened there, or what has or is happening in
> openstack-land.
>
> I would also think it would be great for tenants to be able to setup
> connections into AWS, and even start up AWS instances via the
> openstack API. :)
>
> Thanks,
> Curtis.
>
>>
>> Consensus is just treat each network island like another remote DC and
>> use normal VPN type stuff to glue them together.
>>
>> ( nod to http://romana.io an interesting looking network and security
>> automation project as a network agnostic alternative to SDN for
>> managing cross cloud policy on whatever networks are available. )
>>
>> -Jon
>>
>> ___
>> OpenStack-operators mailing list
>> OpenStack-operators@lists.openstack.org
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
>
>
> --
> Blog: serverascode.com
>
> ___
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

Re: [Openstack] [OpenStack] [keystone] How to make keystone highly available?

[John van Ommen]

If you'd like to avoid the pain of configuring this manually, Helion
OpenStack has many HA features enabled by default:

http://docs.hpcloud.com/#3.x/helion/planning/high_availability.html

On Mon, Sep 19, 2016 at 2:10 PM, Alexandr Porunov
 wrote:
> Thank you for pointing on it. I haven't yet use pacemaker but I will try to
> configure it.
>
> Sincerely,
> Alexandr
>
> On Mon, Sep 19, 2016 at 10:23 PM, Remo Mattei  wrote:
>>
>> There is already ha options with pacemaker and corosync. Have you looked
>> at that?
>>
>> Inviato da iPhone
>>
>> Il giorno 19 set 2016, alle ore 11:46, Alexandr Porunov
>>  ha scritto:
>>
>> Hello,
>>
>> I am thinking about using the keystone as an authentication system but I
>> am afraid about failures which can affect all the cluster. In fact if the
>> keystone server dies then our full cluster will stop. It would be better if
>> we could use HA with the keystone. Then if our primary keystone server dies
>> we have to elect a new primary keystoe server. Are there some tools which
>> can be used in HA deployment?
>>
>> Any piece of advice will be valuable
>>
>> Sincerely,
>> Alexandr
>> !DSPAM:1,57e036cd282281177211101!
>>
>> ___
>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>
>>
>> !DSPAM:1,57e036cd282281177211101!
>
>
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack-operators] OpenStack Consult Opp (Network/Horizon/VNC)

[John van Ommen]

Hewlett Packard Enterprise has a professional services team dedicated to
fulfilling projects like this. We've been doing this for years now, with
numerous clients, all over the world.

I can connect you with a sales rep if you'd like. (I'm on the integration
side, not sales)

John

On Sep 3, 2016 7:40 PM, "Adam Lawson"  wrote:

> Is anyone *against* creating a mailing-list this sort of dialog? It's for
> short-term immediate need/troubleshooting/my hair is on fire soert s of
> requests.
>
> Also, if i wanted to push this forward, what the next step since I'm not
> seeing anyone volunteering other than JJ who tried and that's the last I
> heard of it.
>
> //adam
>
>
> *Adam Lawson*
>
> AQORN, Inc.
> 427 North Tatnall Street
> Ste. 58461
> Wilmington, Delaware 19801-2230
> Toll-free: (844) 4-AQORN-NOW ext. 101
> International: +1 302-387-4660
> Direct: +1 916-246-2072
>
> On Mon, Feb 8, 2016 at 5:20 PM, Robert Starmer  wrote:
>
>> I thought this got stuck in the "do we need another list" and "well, what
>> is our alternative" discussion.  So, no I don't recall any progress.  I
>> still think it'd be useful to have a list. for this class of discussion.
>>
>> Robert
>>
>> On Wed, Feb 3, 2016 at 6:01 PM, Adam Lawson  wrote:
>>
>>> Hey all,
>>>
>>> Just curious how this was progressing. Is there an approval waiting to
>>> happen or something in the background?
>>>
>>> //adam
>>>
>>>
>>> *Adam Lawson*
>>>
>>> AQORN, Inc.
>>> 427 North Tatnall Street
>>> Ste. 58461
>>> Wilmington, Delaware 19801-2230
>>> Toll-free: (844) 4-AQORN-NOW ext. 101
>>> International: +1 302-387-4660
>>> Direct: +1 916-246-2072
>>>
>>> On Tue, Nov 24, 2015 at 1:55 PM, JJ Asghar  wrote:
>>>
 -BEGIN PGP SIGNED MESSAGE-
 Hash: SHA512

 On 11/24/15 3:13 PM, Adam Lawson wrote:
 > Yep. I knew I was walking a gray line. If I had time to let folks
 > know about an opportunity and wait for folks to visit and reply I
 > totally would. Otherwise, I would definitely echo a job-related
 > mailing list if that could be setup?
 >
 >
 > On Tue, Nov 24, 2015 at 12:51 PM, JJ Asghar  > wrote:
 >
 >
 > As another place, other than the job posting site I just created
 > this review[1].
 >
 > If you like this idea please +1 it.
 >
 >
 > [1]: https://review.openstack.org/249415


 Yep, that's that review above. Go ahead and +1 it also, and we'll see
 if we can the list put together. :D


 - --
 Best Regards,
 JJ Asghar
 c: 512.619.0722 t: @jjasghar irc: j^2
 -BEGIN PGP SIGNATURE-
 Version: GnuPG/MacGPG2 v2
 Comment: GPGTools - https://gpgtools.org

 iQIcBAEBCgAGBQJWVNy8AAoJEDZbxzMH0+jTpEAP/3qnKoVP/iVKlFmZG7CrhLxC
 TveL3xNyTBGt+cMVQAJGChSH8v0l1+XuUKeIkRBsPCppe7PnPfIqPuCFJKuvCU+v
 gFDUvJbpqSXL+D7vfvc/83GUPa9rgvV//ovicbrvvJIsvl5LMuLpUgBVhZRDWzu2
 Q+VjdRFkHVCAHepQxBiyRav6Cwa5Sng3ODHrjlhKc4Kr2me25B+NDCqPbQz4/h6a
 T/nnHFrvZ0PQcYmNBONRe2WX7IqwJx/igV3+HuHXUtrtpKPRX7V12k5WdEGqbVDn
 LBEwJDYJWFXa8uo/WsbiaQfsO4iP8TZwP8aU/kcEB685HnRVr6Naw19VZwEvp8ZZ
 5+WoT861u1VwWe244MvWgWLkWPVYsZIzCJAfcNmqVTQByc2VnozTUDoYWirjCPfS
 ka5Fj/t2jEh5fr6cR8x/h4yaq3DO345yVL2U9oj0NhczEpNYpFCdg66Huh8sMgca
 BwZmzd50UGTxq1tp5etT+XBeAqxhwd0iqgNZLXsTLQgJZ58giqgTymtgOGoROz1c
 PbvUoUgSgLTzr9PnyEAV09gZvyYr+8goR6RtfbDTWyykSqDrByr8qqef4DnxkyoM
 Y4lzp+vGkMK3RxJcVgM7fUmbIJTo42fpuwDFC39FmoSICJVqzYuNKcM6NlUmzXHE
 GbKDfBTzXrWTkgo1+0uW
 =T6en
 -END PGP SIGNATURE-

>>>
>>>
>>> ___
>>> OpenStack-operators mailing list
>>> OpenStack-operators@lists.openstack.org
>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>>>
>>>
>>
>
> ___
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
>
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

Re: [Openstack] [OpenStack] [Swift] Is it possible to create chunked upload?

[John van Ommen]

The Swift3 plugin for Swift supports multipart upload.

So you could do the following:

1) install the Swift3 plugin
2) use Amazon's s3 API for multi part upload

http://docs.aws.amazon.com/AmazonS3/latest/dev/mpuoverview.html

On Aug 23, 2016 9:16 AM, "Alexandr Porunov" 
wrote:

> Hello Jay,
>
> Unfortunately I don't know python. I am writing on java. Could you explain
> me briefly what I have to use to send chunks in the Swift and then combine
> those chunks? Or maybe we can append bytes to file somehow?
> If I understand correctly I need to store my first chunk as an individual
> object (0 - 4095 bytes) and then when I recieve the second chunk (with 4096
> - 8192 bytes) I have to append that chunk to the first object. Is it
> correct? How can we append second chunk?
>
> Sincerely,
> Alexandr
>
> On Tue, Aug 23, 2016 at 5:49 PM, Jay Pipes  wrote:
>
>> On 08/23/2016 10:13 AM, Alexandr Porunov wrote:
>>
>>> Hello,
>>>
>>> My server accepts files in chunks (4 Kbytes each chunk. File's size can
>>> be till 8 GB). Is it possible somehow store those chunks in Swift like a
>>> single file? Does somebody know any solution to solve this problem?
>>>
>>
>> Yes, you can do this.
>>
>> The Glance Swift driver shows an example of how to do this:
>>
>> https://github.com/openstack/glance_store/blob/master/glance
>> _store/_drivers/swift/store.py#L829-L944
>>
>> Best,
>> -jay
>>
>> ___
>> Mailing list: http://lists.openstack.org/cgi
>> -bin/mailman/listinfo/openstack
>> Post to : openstack@lists.openstack.org
>> Unsubscribe : http://lists.openstack.org/cgi
>> -bin/mailman/listinfo/openstack
>>
>
>
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/
> openstack
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] S3 API with Swift

[John van Ommen]

I managed to get half of it working. There's an emulation layer
between EC2 and Keystone, and an emulation layer between S3 and Swift.
The first half is working for me now.

The 'trick' was realizing that the Helion distribution deploys the
various OpenStack services in a Python virtual environment.

So, the bad news is that you can't just follow the docs from
OpenStack.org to deploy the S3 emulation layer.
But the *good* news is that you don't have to work on each and every
proxy server. Basically you build a new Python virtual environment,
and the Helion installer deploys that environment to all the proxies.

This is very nice if you're managing a lot of nodes, it makes it
easier to manage. And TBH, it wasn't that difficult to set up once I
realized how it had to be done, I managed to get it working on the
first try once I found the documentation. (Which wasn't so easy.)

Here's a doc on how to do this in Helion, this is how I got it
working: 
https://docs.hpcloud.com/#3.x/helion/operations/thirdparty_driver_support.html

On Fri, Aug 19, 2016 at 3:06 PM, Mark Kirkwood
<mark.kirkw...@catalyst.net.nz> wrote:
> Hi John,
>
> Probably the first thing to do is get a reasonable error display. The
> service init scripts sometime throw away the output so try starting the
> proxy with the command:
>
> $ sudo /usr/bin/swift-init proxy-server start
>
> (or whatever variant your platform uses - see your system's service init
> file - in my case /etc/init.d/swift-proxy for details)
>
> On that note, would be helpful to know your platform and Swift version as
> well.
>
> Cheers
>
> Mark
>
> On 20/08/16 04:48, John van Ommen wrote:
>>
>> Has anyone been able to successfully integrate the S3 api with Swift?
>>
>> I'm working on this in my lab, and finding a number of issues:
>>
>> 1) I've found that when I enable the S3 API, my swift proxy doesn't
>> bind to it's port. I'm guessing that it's failing to start for some
>> reason, but the logs don't say why it's not.
>> 2) I've found that when I disable the S3 API, my swift proxy works
>> fine. So I don't think the problem is Swift, I think the problem is
>> the S3 API.
>> 3) The documentation is very limited. There is a sample configuration
>> file, but little explanation of how things are set or why.
>>
>> It looks like the folks at SwiftStack got this working, as it's
>> featured in their product. IBM also has a page on this, but IBM's
>> configuration doesn't match what's documented at OpenStack.org.
>>
>> I've tried about two dozen combinations, and every time S3 is enabled,
>> Swift won't listen. Any ideas?
>>
>> Here's my config. Note that the stanzas related to S3 are edited out,
>> because I couldn't get Swift to run with them turned on.
>>
>> [DEFAULT]
>> bind_port = 8080
>> bind_ip = 172.16.15.10
>> swift_dir = /etc/swift
>> user = swift
>>
>> [pipeline:main]
>> pipeline = catch_errors gatekeeper healthcheck proxy-logging cache
>> container_sync bulk crossdomain tempurl formpost ratelimit authtoken
>> keystoneauth staticweb container-quotas account-quotas slo dlo
>> versioned_writes proxy-logging name_check proxy-server
>>
>> [app:proxy-server]
>> use = egg:swift#proxy
>> account_autocreate = true
>> sorting_method = timing
>>
>> [filter:catch_errors]
>> use = egg:swift#catch_errors
>>
>> [filter:gatekeeper]
>> use = egg:swift#gatekeeper
>>
>> [filter:healthcheck]
>> use = egg:swift#healthcheck
>>
>> [filter:proxy-logging]
>> use = egg:swift#proxy_logging
>>
>> [filter:cache]
>> use = egg:swift#memcache
>>
>> [filter:container_sync]
>> use = egg:swift#container_sync
>>
>> [filter:bulk]
>> use = egg:swift#bulk
>>
>> [filter:crossdomain]
>> use = egg:swift#crossdomain
>>
>> [filter:tempurl]
>> use = egg:swift#tempurl
>>
>> [filter:formpost]
>> use = egg:swift#formpost
>>
>> [filter:ratelimit]
>> use = egg:swift#ratelimit
>>
>> [filter:authtoken]
>> paste.filter_factory = keystonemiddleware.auth_token:filter_factory
>> identity_uri = https://helion-cp1-vip-KEY-API-mgmt:5000
>> admin_tenant_name = services
>> admin_user = swift
>> admin_password = somepassword
>> auth_uri = https://helion-cp1-vip-KEY-API-mgmt:5000
>> cache = swift.cache
>> include_service_catalog = False
>> delay_auth_decision = true
>> #memcache_security_strategy = ENCRYPT
>> #memcache_secret_key = somekey
>>
>> # Note to reviewer: I'm including all possible filters but not all are
>> # included in the pip

[Openstack] S3 API with Swift

[John van Ommen]

Has anyone been able to successfully integrate the S3 api with Swift?

I'm working on this in my lab, and finding a number of issues:

1) I've found that when I enable the S3 API, my swift proxy doesn't
bind to it's port. I'm guessing that it's failing to start for some
reason, but the logs don't say why it's not.
2) I've found that when I disable the S3 API, my swift proxy works
fine. So I don't think the problem is Swift, I think the problem is
the S3 API.
3) The documentation is very limited. There is a sample configuration
file, but little explanation of how things are set or why.

It looks like the folks at SwiftStack got this working, as it's
featured in their product. IBM also has a page on this, but IBM's
configuration doesn't match what's documented at OpenStack.org.

I've tried about two dozen combinations, and every time S3 is enabled,
Swift won't listen. Any ideas?

Here's my config. Note that the stanzas related to S3 are edited out,
because I couldn't get Swift to run with them turned on.

[DEFAULT]
bind_port = 8080
bind_ip = 172.16.15.10
swift_dir = /etc/swift
user = swift

[pipeline:main]
pipeline = catch_errors gatekeeper healthcheck proxy-logging cache
container_sync bulk crossdomain tempurl formpost ratelimit authtoken
keystoneauth staticweb container-quotas account-quotas slo dlo
versioned_writes proxy-logging name_check proxy-server

[app:proxy-server]
use = egg:swift#proxy
account_autocreate = true
sorting_method = timing

[filter:catch_errors]
use = egg:swift#catch_errors

[filter:gatekeeper]
use = egg:swift#gatekeeper

[filter:healthcheck]
use = egg:swift#healthcheck

[filter:proxy-logging]
use = egg:swift#proxy_logging

[filter:cache]
use = egg:swift#memcache

[filter:container_sync]
use = egg:swift#container_sync

[filter:bulk]
use = egg:swift#bulk

[filter:crossdomain]
use = egg:swift#crossdomain

[filter:tempurl]
use = egg:swift#tempurl

[filter:formpost]
use = egg:swift#formpost

[filter:ratelimit]
use = egg:swift#ratelimit

[filter:authtoken]
paste.filter_factory = keystonemiddleware.auth_token:filter_factory
identity_uri = https://helion-cp1-vip-KEY-API-mgmt:5000
admin_tenant_name = services
admin_user = swift
admin_password = somepassword
auth_uri = https://helion-cp1-vip-KEY-API-mgmt:5000
cache = swift.cache
include_service_catalog = False
delay_auth_decision = true
#memcache_security_strategy = ENCRYPT
#memcache_secret_key = somekey

# Note to reviewer: I'm including all possible filters but not all are
# included in the pipeline (because they are not required)
[filter:versioned_writes]
use = egg:swift#versioned_writes
allow_versioned_writes = true

[filter:keystoneauth]
use = egg:swift#keystoneauth
operator_roles = admin, swiftoperator, _member_, Member

[filter:staticweb]
use = egg:swift#staticweb

[filter:container-quotas]
use = egg:swift#container_quotas

[filter:account-quotas]
use = egg:swift#account_quotas

[filter:slo]
use = egg:swift#slo

# AWS S3 document says that each part must be at least 5 MB in a multipart
# upload, except the last part.
#min_segment_size = 5242880

[filter:dlo]
use = egg:swift#dlo

[filter:domain_remap]
use = egg:swift#domain_remap

[filter:cname_lookup]
use = egg:swift#cname_lookup

[filter:name_check]
use = egg:swift#name_check
forbidden_chars = "<>
maximum_length = 255

[filter:list-endpoints]
use = egg:swift#list_endpoints

[filter:xprofile]
use = egg:swift#xprofile

#[filter:swift3]
#use = egg:swift3#swift3

# Swift has no concept of the S3's resource owner; the resources
# (i.e. containers and objects) created via the Swift API have no owner
# information. This option specifies how the swift3 middleware handles them
# with the S3 API.  If this option is 'false', such kinds of resources will be
# invisible and no users can access them with the S3 API.  If set to 'true',
# the resource without owner is belong to everyone and everyone can access it
# with the S3 API.  If you care about S3 compatibility, set 'false' here.  This
# option makes sense only when the s3_acl option is set to 'true' and your
# Swift cluster has the resources created via the Swift API.
# allow_no_owner = false
#
# Set a region name of your Swift cluster.  Note that Swift3 doesn't choose a
# region of the newly created bucket actually.  This value is used for the
# GET Bucket location API and v4 signatures calculation.
# location = US
#
# Set whether to enforce DNS-compliant bucket names. Note that S3 enforces
# these conventions in all regions except the US Standard region.
# dns_compliant_bucket_names = True
#
# Set the default maximum number of objects returned in the GET Bucket
# response.
# max_bucket_listing = 1000
#
# Set the maximum number of parts returned in the List Parts operation.
# (default: 1000)
# When setting it to be larger than 1, set to be larger
# container_listing_limit in swift.conf.(specification of S3: 1000)
# max_parts_listing = 1000
#
# Set the maximum number of objects we can delete with the Multi-Object Delete
# operation.
# 

Re: [Openstack-operators] Who's using TripleO in production?

[John van Ommen]

The 1.x versions of Helion Open Stack used Triple O.

On Aug 2, 2016 9:59 AM, "Curtis"  wrote:

> Hi,
>
> I'm just curious who, if anyone, is using TripleO in production?
>
> I'm having a hard time finding anywhere to ask end-user type
> questions. #tripleo irc seems to be just a dev channel. Not sure if
> there is anywhere for end users to ask questions. A quick look at
> stackalytics shows it's mostly RedHat contributions, though again, it
> was a quick look.
>
> If there were other users it would be cool to perhaps try to have a
> session on it at the upcoming ops midcycle.
>
> Thanks,
> Curtis.
>
> ___
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

[Openstack] dumb neutron question

[John van Ommen]

Let's say I have two networks. One network is 10.241.0.1/24, and my
VMs are on that network. The other network is a provider network with
a CIDR of 192.168.100.0/22.

There's an illustration of what I mean here:

https://developer.rackspace.com/blog/neutron-networking-l3-agent/

Okay, so here's my question:
There's a router between the two networks. Is the routing being done
by a *physical* router, or by a *neutron* router?


Here's the reason that I ask:
When I set these networks up, I generally set up a neutron router
between the first network and the second network. But in my current
project, there's a physical router at 192.168.100.1. So it seems like
my Neutron router would be in conflict with the physical router at
192.168.100.1.

IE, I can't have two routers running on the same subnet. Or am I
missing something?

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] Multi-Node OpenStack with no VLANs

[John van Ommen]

I am working on an OpenStack deployment where my customer's VLANs do
not appeared to be configured properly.

I am in a time crunch and I'm looking for a way to work around this.

Would it be possible to use VxLAN for my guest traffic, and basically
pick the subnets arbitrarily?

I am using OpenStack Kilo with KVM for my compute nodes, and I'm using
OVS for switching. It *seems* like I could pick the networks "out of
thin air" with VxLAN, and let OVS deal with the tags, instead of using
VLAN and have the customer's switches deal with the tags.

Will that work?

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack-operators] Triple-O: number of concurrent installations

[John van Ommen]

Excellent, thanks for following up.

On Tue, Mar 1, 2016 at 7:15 AM, Uwe Sauter <uwe.sauter...@gmail.com> wrote:

> Hi,
>
> I changed the setting on Undercloud's Nova and can confirm that it indeed
> is what I was looking for.
>
>
> Thanks again,
>
> Uwe
>
>
> Am 01.03.2016 um 14:03 schrieb John van Ommen:
> > Mr. Simard would know better than I, as I'm more familiar with Helion's
> > Triple-O versus RDO's Triple-O.
> >
> > Having said that, yes, the Undercloud provisions the Overcloud and
> > therefore making the change in "nova.conf" prior to provisioning the
> > Overcloud would accomplish what you're looking for.
> >
> > On Tue, Mar 1, 2016 at 1:52 AM, Uwe Sauter <uwe.sauter...@gmail.com>
> wrote:
> >
> >> Hi David,
> >>
> >> as I'm not that familiar with the internals of Triple-O: this is a
> setting
> >> that I would change on the Undercloud host ("director"
> >> in Red Hat terminology)?
> >>
> >> Thank you,
> >>
> >> Uwe
> >>
> >>
> >>
> >>
> >> Am 29.02.2016 um 23:37 schrieb David Moreau Simard:
> >>> Hi Uwe,
> >>>
> >>> It looks like the configuration option you're looking for is
> >>> max_concurrent_builds from nova.conf [1].
> >>>
> >>> Feel free to reach out on #tripleo on RDO if you have any questions!
> >>>
> >>> [1]:
> >>
> http://docs.openstack.org/liberty/config-reference/content/list-of-compute-config-options.html
> >>>
> >>> David Moreau Simard
> >>> Senior Software Engineer | Openstack RDO
> >>>
> >>> dmsimard = [irc, github, twitter]
> >>>
> >>>
> >>> On Mon, Feb 29, 2016 at 10:52 AM, Uwe Sauter <uwe.sauter...@gmail.com>
> >> wrote:
> >>>> Hi all,
> >>>>
> >>>> I'm searching for a setting in Triple-O where I can limit the number
> of
> >> concurrent / parallel installations. I'm facing the issue
> >>>> that deployments hang or even fail due to bandwidth limit in the
> >> provisioning network. I think that I can work around this issue
> >>>> if I can limit the number of nodes that boot up in parallel.
> >>>>
> >>>> Has someone already done this? Which setting do I need to tweak for
> >> this?
> >>>>
> >>>>
> >>>> Thanks in advance,
> >>>>
> >>>> Uwe
> >>>>
> >>>> ___
> >>>> OpenStack-operators mailing list
> >>>> OpenStack-operators@lists.openstack.org
> >>>>
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
> >>
> >>
> >> ___
> >> OpenStack-operators mailing list
> >> OpenStack-operators@lists.openstack.org
> >> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
> >>
> >
>
>
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

Re: [Openstack-operators] Triple-O: number of concurrent installations

[John van Ommen]

Mr. Simard would know better than I, as I'm more familiar with Helion's
Triple-O versus RDO's Triple-O.

Having said that, yes, the Undercloud provisions the Overcloud and
therefore making the change in "nova.conf" prior to provisioning the
Overcloud would accomplish what you're looking for.

On Tue, Mar 1, 2016 at 1:52 AM, Uwe Sauter  wrote:

> Hi David,
>
> as I'm not that familiar with the internals of Triple-O: this is a setting
> that I would change on the Undercloud host ("director"
> in Red Hat terminology)?
>
> Thank you,
>
> Uwe
>
>
>
>
> Am 29.02.2016 um 23:37 schrieb David Moreau Simard:
> > Hi Uwe,
> >
> > It looks like the configuration option you're looking for is
> > max_concurrent_builds from nova.conf [1].
> >
> > Feel free to reach out on #tripleo on RDO if you have any questions!
> >
> > [1]:
> http://docs.openstack.org/liberty/config-reference/content/list-of-compute-config-options.html
> >
> > David Moreau Simard
> > Senior Software Engineer | Openstack RDO
> >
> > dmsimard = [irc, github, twitter]
> >
> >
> > On Mon, Feb 29, 2016 at 10:52 AM, Uwe Sauter 
> wrote:
> >> Hi all,
> >>
> >> I'm searching for a setting in Triple-O where I can limit the number of
> concurrent / parallel installations. I'm facing the issue
> >> that deployments hang or even fail due to bandwidth limit in the
> provisioning network. I think that I can work around this issue
> >> if I can limit the number of nodes that boot up in parallel.
> >>
> >> Has someone already done this? Which setting do I need to tweak for
> this?
> >>
> >>
> >> Thanks in advance,
> >>
> >> Uwe
> >>
> >> ___
> >> OpenStack-operators mailing list
> >> OpenStack-operators@lists.openstack.org
> >> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
>
> ___
> OpenStack-operators mailing list
> OpenStack-operators@lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators

Re: [Openstack] One Interface with VLAN and one with VxLAN

[John van Ommen]

Does that depend on DVR to function?

As I understand it, OpenStack Kilo uses DVR by default when you use KVM for
your compute.
But with vCenter for compute, I have a hunch that I may need NSX to make
this work. (Because there's not distributed virtual router for the ovsvapp
nodes.)

But, again, I could be wrong on this.

On Thu, Feb 25, 2016 at 6:57 PM, Han Chao <hanc...@unitedstack.com> wrote:

> Hi,
>
> If you would have both VLAN and VxLAN types of provider networks, it is
> possible to accomplish that.
>
> First of all, you need to provide the right configuration in the file of
> "openvswitch_agent.ini" and "ml2_conf.ini". Then with two NICs (ports),
> connect them to the different bridges (one for VLAN (br-vlan) and the other
> for VxLAN (br-tun)), respectively. At this stage, you should have two types
> of provider networks, one for VLAN and the other for VxLAN.
>
> Br,
> Chao
>
> -- Original --
> *From: * "John van Ommen"<john.vanom...@gmail.com>;
> *Date: * Fri, Feb 26, 2016 03:46 AM
> *To: * "Neil Jerram"<neil.jer...@metaswitch.com>;
> *Cc: * "openstack@lists.openstack.org"<openstack@lists.openstack.org>;
> *Subject: * Re: [Openstack] One Interface with VLAN and one with VxLAN
>
> I want to have VLAN provider networks for some customers.
>
> I currently have my cloud configured for VxLAN, and I'm not able to create
> a VLAN provider network.
>
> On Thu, Feb 25, 2016 at 11:24 AM, Neil Jerram <neil.jer...@metaswitch.com>
> wrote:
>
>> On 25/02/16 19:08, John van Ommen wrote:
>> > Is it possible to have one interface configured with VLAN and another
>> > with VxLAN?
>>
>> Do you mean for tenant networks, or for provider networks?
>>
>> I believe it's possible for provider networks, but not for tenant. [1]
>>
>> [1]
>>
>> http://lists.openstack.org/pipermail/openstack-dev/2016-January/084744.html
>>
>> > I'm in a situation where I have some customers who want to do VxLAN, but
>> > I also have some legacy customers that need VLAN.
>>
>> Out of interest, why do the legacy customers need VLAN specifically?
>>
>> > To complicate matters further, my compute nodes are in vCenter. I
>> > understand that KVM can do VxLAN and VLAN using DVR. My vCenter compute
>> > nodes are plugged into a OVSVapp virtual switch.
>>
>> I'm afraid I have no idea about this dimension!
>>
>> Regards,
>> Neil
>>
>>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] One Interface with VLAN and one with VxLAN

[John van Ommen]

I want to have VLAN provider networks for some customers.

I currently have my cloud configured for VxLAN, and I'm not able to create
a VLAN provider network.

On Thu, Feb 25, 2016 at 11:24 AM, Neil Jerram <neil.jer...@metaswitch.com>
wrote:

> On 25/02/16 19:08, John van Ommen wrote:
> > Is it possible to have one interface configured with VLAN and another
> > with VxLAN?
>
> Do you mean for tenant networks, or for provider networks?
>
> I believe it's possible for provider networks, but not for tenant. [1]
>
> [1]
> http://lists.openstack.org/pipermail/openstack-dev/2016-January/084744.html
>
> > I'm in a situation where I have some customers who want to do VxLAN, but
> > I also have some legacy customers that need VLAN.
>
> Out of interest, why do the legacy customers need VLAN specifically?
>
> > To complicate matters further, my compute nodes are in vCenter. I
> > understand that KVM can do VxLAN and VLAN using DVR. My vCenter compute
> > nodes are plugged into a OVSVapp virtual switch.
>
> I'm afraid I have no idea about this dimension!
>
> Regards,
> Neil
>
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] One Interface with VLAN and one with VxLAN

[John van Ommen]

Is it possible to have one interface configured with VLAN and another with
VxLAN?

I'm in a situation where I have some customers who want to do VxLAN, but I
also have some legacy customers that need VLAN.

To complicate matters further, my compute nodes are in vCenter. I
understand that KVM can do VxLAN and VLAN using DVR. My vCenter compute
nodes are plugged into a OVSVapp virtual switch.
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] CEPH Speed Limit

[John van Ommen]

I have a client who isn't happy with the performance of their storage.
The client is currently running a mix of SAS HDDs and SATA SSDs.

They wanted to remove the SAS HDDs and replace them with SSDs, so the
entire array would be SSDs.

I was running benchmarks on the current hardware and I found that the
performance of the HDD array was close to the performance of the SSD array.

To me, this indicates that we're reaching the limits of the controller that
it's attached to. (An LSI RAID controller that's built into the system
board.)

I was about to recommend that they add a controller, when I realized that
we may be reaching the limits of the PCI-E bus itself.

Before I go and make a bad recommendation, I have a few questions:

1) Am I correct in assuming that the RAID controller, though physically on
the system board, is still running through the PCI-E bus, just as if it was
plugged into a slot?
2) Am I correct in assuming that the limit for the PCI-E bus (version 2) is
500Mb/s? (https://en.wikipedia.org/wiki/PCI_Express)

And if points one and two are correct, is my hypothesis that adding more
SSDs won't improve things true?

Right now my benchmarks are showing that sequential reads are hitting about
600Mb/s. (I haven't confirmed if their server is PCI-E 2.0 or 3.0)

John
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] VLAN Provider Question

[John van Ommen]

Excellent! Thanks for clarifying how this works. I was reading the docs
from James Denton and the one from Openstack.org and getting quite confused.

On Fri, Jan 15, 2016 at 8:17 AM, gustavo panizzo (gfa) <g...@zumbi.com.ar>
wrote:

> On Fri, Jan 15, 2016 at 07:38:37 -0800, John van Ommen wrote:
> > My compute node and my controller node are both physically connected to
> the
> > same switch, and they can talk to the same tagged VLAN.
> >
> > The part I'm having trouble understanding is the network namespace. For
> > instance:
> >
> > I have a provider VLAN on VLAN 100
> > The CIDR for VLAN 100 is 192.168.0.0/24
> >
> > Is there an IP address on VLAN 100 for the physical interface? Or does
> that
> > IP range only exist in the network namespace? I understand that the DHCP
> > server listens on the network namespace. I'm trying to understand what
> must
> > exist OUTSIDE of that network namespace, on the physical interface that
> > hosts it (eth0).
>
> the dhcp server port will be created on a network namespace, it will have
> an ip on 192.168.0.0/24 network. Your physical interface eth0, and the
> vswitch that holds it, won't have an ip on that network
>
> >
> > On Fri, Jan 15, 2016 at 5:00 AM, Andreas Scheuring <
> > scheu...@linux.vnet.ibm.com> wrote:
> >
> > > Hi John,
> > > the short answer: If you dhcp runs on the network node, then of course
> > > your network node must be attached to that provider network as well!
> > >
> > > Andreas (IRC: scheuran)
> > >
> > > On Do, 2016-01-14 at 22:08 -0800, John van Ommen wrote:
> > > > If I have a VM that's getting it's IP addresses from a VLAN provider
> > > > network, is it sufficient to have a network namespace that
> corresponds
> > > > to that network?
> > > >
> > > > Or do I also need to have a network interface on the server that
> > > > corresponds to it?
> > > >
> > > > For instance, if I have a provider network with a CIDR of
> > > > 192.168.0.1/24, and I have a network namespace that maps to that
> > > > provider network, is that sufficient? Or do I also need to have a
> > > > network interface on the control plane and network node in the same
> > > > CIDR?
> > > > ___
> > > > Mailing list:
> > > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> > > > Post to : openstack@lists.openstack.org
> > > > Unsubscribe :
> > > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> > >
> > >
> > >
>
> > ___
> > Mailing list:
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> > Post to : openstack@lists.openstack.org
> > Unsubscribe :
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>
>
> --
> 1AE0 322E B8F7 4717 BDEA BF1D 44BB 1BA7 9F6C 6333
>
> keybase: http://keybase.io/gfa
>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack] VLAN Provider Question

[John van Ommen]

If I have a VM that's getting it's IP addresses from a VLAN provider
network, is it sufficient to have a network namespace that corresponds to
that network?

Or do I also need to have a network interface on the server that
corresponds to it?

For instance, if I have a provider network with a CIDR of 192.168.0.1/24,
and I have a network namespace that maps to that provider network, is that
sufficient? Or do I also need to have a network interface on the control
plane and network node in the same CIDR?
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

[Openstack-operators] Attempting to Reserve Node

[John van Ommen]

I'm trying to provision nodes with Ironic, and my Ironic logs show that
Ironic is attempting to reserve node.

It continues to do this forever.

I took a look at the code and it appears to be some type of lock condition.

Anyone know the solution?

John
___
OpenStack-operators mailing list
OpenStack-operators@lists.openstack.org
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators