Re: Primary Storage algorithm - when having more primary storages

[Andrija Panic]

THanks guys for this info!

On 21 April 2015 at 01:32, Prachi Damle  wrote:

> By Default it is first fit - and that means the first one found with
> enough capacity. This results in the first one returned by the Db ordering
> to get allocated first until it has capacity.
> One can choose the algorithm using vm.allocation.algorithm as Abhi
> mentioned.
>
> -Prachi
>
> -Original Message-
> From: Abhinandan Prateek [mailto:abhinandan.prat...@shapeblue.com]
> Sent: Monday, April 20, 2015 8:37 AM
> To: dev@cloudstack.apache.org
> Subject: Re: Primary Storage algorithm - when having more primary storages
>
> Following algorithms are used to select storage pool:
>
> random, userconcentratedpod_random: RANDOM
> userdispersing: Select on basis of least no of volumes
> firstfitleastconsumed: Select on basis of remaining capacity
>
> These values can be set for “vm.allocation.algorithm” global config
> variable.
>
> -abhi
>
> > On 20-Apr-2015, at 8:17 pm, Andrija Panic 
> wrote:
> >
> > Sorry for my ignorance - meaing completely random (except the scope ?)
> >
> > On 20 April 2015 at 16:01, Wei ZHOU  wrote:
> >
> >> /etc/cloudstack/management/componentContext.xml
> >>
> >>   >> class="com.cloud.utils.component.AdapterList">
> >>
> >>  
> >>  
> >>  
> >>  
> >>  
> >>
> >>  
> >>
> >>
> >> 2015-04-20 15:44 GMT+02:00 Andrija Panic :
> >>
> >>> Hi people,
> >>>
> >>> just wondering - if I have 3 primary storages (3 x NFS boxes) - what is
> >> the
> >>> volume alocation algoritm - is it compelte random, or is it like 1
> >> client's
> >>> volume goes always to 1 storage...
> >>>
> >>> Any info appreciated !
> >>>
> >>> Thanks,
> >>>
> >>> --
> >>>
> >>> Andrija Panić
> >>>
> >>
> >
> >
> >
> > --
> >
> > Andrija Panić
>
> Find out more about ShapeBlue and our range of CloudStack related services
>
> IaaS Cloud Design & Build<
> http://secure-web.cisco.com/1PRv-JvQmbK3oYfgoOToEgShQXolmGgMLxjLUpY0FODAUAMVH4eRiAXTSud4jMt2CqCcHg09yk2UrBoy_Z6caE4aRXiSeTsH2pj26KFr8X1QQjP4o4WRxTWrE0NUVkD0xFN7aXwTXBR0OSXuGYYXHvrKDJ9YRYER_fuh_Qg8zoraTz8M1EtjTh2Og0yzlgahe/http%3A%2F%2Fshapeblue.com%2Fiaas-cloud-design-and-build%2F%2F
> >
> CSForge – rapid IaaS deployment framework<
> http://secure-web.cisco.com/1xqIwQE4oaptduHj2eM0C5Xak_ej-2RW-pw2UnrLy-WxErqQYIcZ-Ks2eBzC9wG5mXkux5dYNWycwQ0em7FHd68FVBp9V_NOjVigLOffysYbM1R4UjuUFm1Nf-ZQc3s7TKGIYaeXzuNim-qXPFQEkotJ-Ej5-ACppDG9xv5CRnbmybT0hWMkbODY-y5gF3zPt/http%3A%2F%2Fshapeblue.com%2Fcsforge%2F
> >
> CloudStack Consulting<
> http://secure-web.cisco.com/1BJ6ln-uVs-Unp3xam7jwZFUOckTBvgsLnG4WEQ6J5LxCozdXLNW7sOmlQ2HU9201euLb-2KzG3yEL7yPM_mUiSW-F6JZziSX_NOB_TDWs86ZpQL92VF8lXBoJc4oEJMgbC1N4fvmpr_v46W3erreM_050Cz4G1wPFuIcan271v6CYFPRXps0jBPLO6CFlikH/http%3A%2F%2Fshapeblue.com%2Fcloudstack-consultancy%2F
> >
> CloudStack Software Engineering<
> http://secure-web.cisco.com/19vJTGBujOo9_SZqEEcXrQuPfPleKldd1EC4Tpolrykx-0aRbmVK5EhvLY9N0GUKpaUVnsaLR8mJ0c-TVjB5sSge55QzFZpkYrAWGq31uT_itILAZUr9Hjijv0X9UVGE2mt2Qey31f-xcJDxB2XK-Iww-MKol6ho683JUbVcLRTPC9eW-u2-Je4jyJbOItt73/http%3A%2F%2Fshapeblue.com%2Fcloudstack-software-engineering%2F
> >
> CloudStack Infrastructure Support<
> http://secure-web.cisco.com/19CvSa6gGGillRrhNixj3SL7LBX-_ARUG2p2IY-Ktd8KDZQ_tV982oGz7MsFpiT2nx2CwpRPtWsFFY5XchJJGe5jrD6SEv-NHf4J1xdX7tgdQiImeGbcqyvRBpNr8XltbfsS-ido_W0GNiNsbMJOeB0XvBUGQSDsDCRSxOPztZHtCM5iGyK2jNsCXZKjQiw_8/http%3A%2F%2Fshapeblue.com%2Fcloudstack-infrastructure-support%2F
> >
> CloudStack Bootcamp Training Courses<
> http://secure-web.cisco.com/1qF9YOVMu2C4EKdpgRLLGDhijeJdMEqGFmtUFdiRVq40mbxoYdeFUqunCEBenuJCEowDhT8JfRJ90cG2Y3M4Nx_485mCuYCJn_yyMgq45-SGvosZ3IRQiSjQETjXaYxj4GR737c8onjv5-TnbWBtUk7_WHJXwu3Qj61lSmi0SQX-qpjZLtQWZvbvuXcO5OC7g/http%3A%2F%2Fshapeblue.com%2Fcloudstack-training%2F
> >
>
> This email and any attachments to it may be confidential and are intended
> solely for the use of the individual to whom it is addressed. Any views or
> opinions expressed are solely those of the author and do not necessarily
> represent those of Shape Blue Ltd or related companies. If you are not the
> intended recipient of this email, you must neither take any action based
> upon its contents, nor copy or show it to anyone. Please contact the sender
> if you believe you have received this email in error. Shape Blue Ltd is a
> company incorporated in England & Wales. ShapeBlue Services India LLP is a
> company incorporated in India and is operated under license from Shape Blue
> Ltd. Shape Blue Brasil Consultoria Ltda is a company incorporated in Brasil
> and is operated under license from Shape Blue Ltd. ShapeBlue SA Pty Ltd is
> a company registered by The Republic of South Africa and is traded under
> license from Shape Blue Ltd. ShapeBlue is a registered trademark.
>



-- 

Andrija Panić

release notes for 4.4.3

[Daan Hoogland]

Can someone link me to a description on how to publish the release
notes. I did update the site but the link is broken. I think
Pierre-Luc already prepared them (@Pierre-Luc: did you?)

AFAICT: this is the last step before sending out an announce.

regards,
-- 
Daan

[DISCUSS] [UI] List of timezones available in the picker and potentially other places

[Erik Weber]

When you set up recurring snapshots, and potentially other places in the UI
where you have to relate to time, you choose timezone from a table looking
like [1].

I, and some of our customers, find this very hard to navigate.
The list currently contains 618 items.. 618...

I'd like to propose to shorten this list, perhaps to a few common
abbreviations only (PDT/EST/GMT/CET etc.), but would like feedback from
others first.



[1]
https://www.dropbox.com/s/iocoek507q0c9r8/Screenshot%202015-04-21%2010.10.21.png?dl=0

-- 
Erik

Re: [DISCUSS] [UI] List of timezones available in the picker and potentially other places

[Milamber]

I think a simply alphabetical order would be sufficient. Or perhaps with
a 2 steps, first choose the continent, second choose the city.


On 21/04/2015 09:21, Erik Weber wrote:
> When you set up recurring snapshots, and potentially other places in the UI
> where you have to relate to time, you choose timezone from a table looking
> like [1].
>
> I, and some of our customers, find this very hard to navigate.
> The list currently contains 618 items.. 618...
>
> I'd like to propose to shorten this list, perhaps to a few common
> abbreviations only (PDT/EST/GMT/CET etc.), but would like feedback from
> others first.
>
>
>
> [1]
> https://www.dropbox.com/s/iocoek507q0c9r8/Screenshot%202015-04-21%2010.10.21.png?dl=0
>

[GitHub] cloudstack pull request:

[abhinavroy02]

Github user abhinavroy02 commented on the pull request:


https://github.com/apache/cloudstack/commit/5ba7689e37362ed11662755205d8e8ae20744a1e#commitcomment-10821109
  
Incorporated the review comments 
1. Check for VM accessibility by logging into VMs after migration.
2. used userapiclient for user operations like deploy VM, attach, detach, 
restore etc
3. Test for systemvm state when host/storage maintenance is done.
4. Check for available host capacity before migrating a VM.
5. In the teardown section added a check to cencel host/storage maintenance 
if there is any.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

Re: [DISCUSS] [UI] List of timezones available in the picker and potentially other places

[Erik Weber]

I'm not sure I agree

My main issue is the number of elements, not necessarily their order.
Now matter how you order ~600 elements it'll be a mess to find the right one

There's quite a lot of redundancy here, for my local timezone (UTC+1 / CET)
there are multiple options who all maps to the same timezone:
- Etc/GMT+1
- Europe/Oslo (or any other city in CET)
- CET

If I'd been in america there'd even be a couple of of SystemV abbreviations
available.
My question is; do we really need all this? We're asking for timezone, not
location after all.

If we are to keep the list, would it make sense to have a condensed version
on top, showing all timezones from Etc ?


--
Erik

On Tue, Apr 21, 2015 at 11:07 AM, Milamber  wrote:

>
> I think a simply alphabetical order would be sufficient. Or perhaps with
> a 2 steps, first choose the continent, second choose the city.
>
>
> On 21/04/2015 09:21, Erik Weber wrote:
> > When you set up recurring snapshots, and potentially other places in the
> UI
> > where you have to relate to time, you choose timezone from a table
> looking
> > like [1].
> >
> > I, and some of our customers, find this very hard to navigate.
> > The list currently contains 618 items.. 618...
> >
> > I'd like to propose to shorten this list, perhaps to a few common
> > abbreviations only (PDT/EST/GMT/CET etc.), but would like feedback from
> > others first.
> >
> >
> >
> > [1]
> >
> https://www.dropbox.com/s/iocoek507q0c9r8/Screenshot%202015-04-21%2010.10.21.png?dl=0
> >
>
>

Next ACS release?

[Andrei Mikhailovsky]

Hello guys, 

Looking at the dev and user lists it is becoming less certain if version 4.5.x 
is ever coming out. It seems like a few months have passed since the not so 
fortunate release of 4.5.0 and I can't find a release schedule for the 4.5.1, 
which seems to have stopped at rc2 stage and haven't progressed further to a 
release stage. 

Are we likely to see any progress with the 4.5.x branch or is the community 
switching towards the 4.6.x branch without releasing the 4.5.x? 

I am a bit unclear as there are no release dates, schedules or dead lines that 
the community should work with. Possibly as a result of this, the ACS releases 
are not being released on time or fast enough. 

Does it make sense to introduce release schedules for ACS that the dev 
community should stick to? Similar to what is being done in many other 
projects, like Ubuntu, etc. Or would this break the ACS project releases even 
more? 

Andrei 

RE: [DISCUSS] [UI] List of timezones available in the picker and potentially other places

[Stephen Turner]

The Etc ones won't work as they don't obey DST.

The ones on your picture seem to be the standard tzconfig timezones. If we want 
to produce a smaller list, we'd have to make a mapping of the small list onto 
standard timezone rules, and maintain it whenever the rules change. (E.g. when 
countries split or merge or decide to change to a different timezone).

-- 
Stephen Turner


-Original Message-
From: Erik Weber [mailto:terbol...@gmail.com] 
Sent: 21 April 2015 11:03
To: dev
Subject: Re: [DISCUSS] [UI] List of timezones available in the picker and 
potentially other places

I'm not sure I agree

My main issue is the number of elements, not necessarily their order.
Now matter how you order ~600 elements it'll be a mess to find the right one

There's quite a lot of redundancy here, for my local timezone (UTC+1 / CET) 
there are multiple options who all maps to the same timezone:
- Etc/GMT+1
- Europe/Oslo (or any other city in CET)
- CET

If I'd been in america there'd even be a couple of of SystemV abbreviations 
available.
My question is; do we really need all this? We're asking for timezone, not 
location after all.

If we are to keep the list, would it make sense to have a condensed version on 
top, showing all timezones from Etc ?


--
Erik

On Tue, Apr 21, 2015 at 11:07 AM, Milamber  wrote:

>
> I think a simply alphabetical order would be sufficient. Or perhaps 
> with a 2 steps, first choose the continent, second choose the city.
>
>
> On 21/04/2015 09:21, Erik Weber wrote:
> > When you set up recurring snapshots, and potentially other places in 
> > the
> UI
> > where you have to relate to time, you choose timezone from a table
> looking
> > like [1].
> >
> > I, and some of our customers, find this very hard to navigate.
> > The list currently contains 618 items.. 618...
> >
> > I'd like to propose to shorten this list, perhaps to a few common 
> > abbreviations only (PDT/EST/GMT/CET etc.), but would like feedback 
> > from others first.
> >
> >
> >
> > [1]
> >
> https://www.dropbox.com/s/iocoek507q0c9r8/Screenshot%202015-04-21%2010
> .10.21.png?dl=0
> >
>
>

Re: Next ACS release?

[Daan Hoogland]

Andrei, we just released 4.4.3 save the release notes and hence the
announcement. Full focus should be on stabalizing and finalizing
4.5.1, now. We didn't discuss going forth with 4.6 yet, though the
topic of release management for 4.6 has passed the dev list. Are there
any concerns you have with the 4.5 release in particular?

The idea of abandoning 4.5 in favour of 4.6 does not appeal to me
much. We at our shop will probably be skipping
4.5.. I can not say much on release schedule/dates. I think
making fixed promises makes only sense if we can easily abandon
features that are proven to be unstable. At the moment our release
process doesn't facilitate this. So the answer to your last question
is yes. We are working to move away from this impediment and will
probably vote for fixed short release cycles in the future where fixes
go in 2 to 4 week cycles and features in 2 to 4 month cycles.

Hope you agree. If not, this discussion is still open.

On Tue, Apr 21, 2015 at 10:26 AM, Andrei Mikhailovsky  wrote:
> Hello guys,
>
> Looking at the dev and user lists it is becoming less certain if version 
> 4.5.x is ever coming out. It seems like a few months have passed since the 
> not so fortunate release of 4.5.0 and I can't find a release schedule for the 
> 4.5.1, which seems to have stopped at rc2 stage and haven't progressed 
> further to a release stage.
>
> Are we likely to see any progress with the 4.5.x branch or is the community 
> switching towards the 4.6.x branch without releasing the 4.5.x?
>
> I am a bit unclear as there are no release dates, schedules or dead lines 
> that the community should work with. Possibly as a result of this, the ACS 
> releases are not being released on time or fast enough.
>
> Does it make sense to introduce release schedules for ACS that the dev 
> community should stick to? Similar to what is being done in many other 
> projects, like Ubuntu, etc. Or would this break the ACS project releases even 
> more?
>
> Andrei



-- 
Daan

[GitHub] cloudstack pull request: Add more verifications to make sure that ...

[sanju1010]

GitHub user sanju1010 opened a pull request:

https://github.com/apache/cloudstack/pull/183

Add more verifications to make sure that removeIpFromNic works fine in case 
of shared network

Added following steps to existing test in case of shared network:
1.After removing secondary ip from nic verify that the allocated state is 
marked as NULL in user_ip_address table for the ip which was released
2.Add the same ip address again to the same nic to make sure that previous 
ip removal was successful
3.Then remove it

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/sanju1010/cloudstack CS-38815

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/cloudstack/pull/183.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #183


commit 08c30576d3b84b695257215f55518d7bbacbd02e
Author: sanjeev 
Date:   2015-04-21T12:11:38Z

Added additional verification steps to make sure that removing secondary ip 
from nic works fine

commit 254c9ecb28fd9a5b8cbeec714fe02ea0e8489010
Author: sanjeev 
Date:   2015-04-21T12:18:44Z

Removed tag from the attr




---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

[GitHub] cloudstack pull request: Escalations

[shwetaag]

GitHub user shwetaag opened a pull request:

https://github.com/apache/cloudstack/pull/184

Escalations

incorporated review comments

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/shwetaag/cloudstack escalations

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/cloudstack/pull/184.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #184


commit 382a171230d4f36a38c6d4518713e87934861748
Author: shweta agarwal 
Date:   2015-04-09T11:21:50Z

verifying template creation permission

commit 3714a4e545bb05984666ab3643995469269cc6ff
Author: shweta agarwal 
Date:   2015-04-21T12:25:56Z

incorporated review comments




---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

[DISCUSS] Moving to Java 8

[Wilder Rodrigues]

Hi all,

Yesterday I started working on the LibvirtComputingResource class in order to 
apply the same patterns I used in the CitrixResourceBase + add more unit tests 
to it After 10 hours of work I got a bit stuck with the 1st test, which would 
cover the refactored LibvirtStopCommandWrapper. Why did I get stuck? The class 
used a few static methods that call native libraries, which I would like to 
mock. However, when writing the tests I faced problems with the current 
Mockito/PowerMock we are using: they are simply not enough for the task.

What did I do then? I added a dependency to EasyMock and PowerMock-EasyMock 
API. It worked almost fine, but I had to add a “-noverify” to both my Eclipse 
Runtime configuration and also to the cloud-plugin-hypervisor-kvm/pom.xml file. 
I agree that’s not nice, but was my first attempt of getting it to work. After 
trying to first full build I faced more problems related to 
ClassDefNotFoundExpcetion which were complaining about Mockito classes. I then 
found out that adding the PowerMockRunner to all the tests classes was going to 
be a heavy burden and would also mess up future changes (e.g. the -noverify 
flag was removed from Java 8, thus adding it now would be a problem soon).

Now that the first 2 paragraphs explain a bit about the problem, let’s get to 
the solution: Java 8

The VerifyError that I was getting was due to the use of the latest EasyMock  
release (3.3.1). I tried to downgrade it to 3.1/3.2 but it also did not work. 
My decision: do not refactor if the proper tests cannot be added. This left me 
with one action: migrate to Java 8.

There were mentions about Java 8 in february[1] and now I will put some energy 
in making it happen.

What is your opinion on it?

Thanks in advance.

Cheers,
Wilder

http://mail-archives.apache.org/mod_mbox/cloudstack-dev/201502.mbox/%3c54eef6be.5040...@shapeblue.com%3E>

Re: [DISCUSS] Moving to Java 8

[Wido den Hollander]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1



On 04/21/2015 03:27 PM, Wilder Rodrigues wrote:
> Hi all,
> 
> Yesterday I started working on the LibvirtComputingResource class
> in order to apply the same patterns I used in the
> CitrixResourceBase + add more unit tests to it After 10 hours of
> work I got a bit stuck with the 1st test, which would cover the
> refactored LibvirtStopCommandWrapper. Why did I get stuck? The
> class used a few static methods that call native libraries, which I
> would like to mock. However, when writing the tests I faced
> problems with the current Mockito/PowerMock we are using: they are
> simply not enough for the task.
> 
> What did I do then? I added a dependency to EasyMock and
> PowerMock-EasyMock API. It worked almost fine, but I had to add a
> “-noverify” to both my Eclipse Runtime configuration and also to
> the cloud-plugin-hypervisor-kvm/pom.xml file. I agree that’s not
> nice, but was my first attempt of getting it to work. After trying
> to first full build I faced more problems related to
> ClassDefNotFoundExpcetion which were complaining about Mockito
> classes. I then found out that adding the PowerMockRunner to all
> the tests classes was going to be a heavy burden and would also
> mess up future changes (e.g. the -noverify flag was removed from
> Java 8, thus adding it now would be a problem soon).
> 
> Now that the first 2 paragraphs explain a bit about the problem,
> let’s get to the solution: Java 8
> 
> The VerifyError that I was getting was due to the use of the latest
> EasyMock  release (3.3.1). I tried to downgrade it to 3.1/3.2 but
> it also did not work. My decision: do not refactor if the proper
> tests cannot be added. This left me with one action: migrate to
> Java 8.
> 
> There were mentions about Java 8 in february[1] and now I will put
> some energy in making it happen.
> 
> What is your opinion on it?
> 

I'm not against it technically, but practically I am.

Ubuntu 14.04 does not ship a Java 8 JRE in the repositories.

CentOS 7 has java-1.8.0-openjdk.x86_64 available, so it would work
there. But Ubuntu is also widely used with CloudStack, so those users
couldn't use CloudStack without any additional repositories.

Since that isn't easy I would vote -1 on this if it came that far.

Wido

> Thanks in advance.
> 
> Cheers, Wilder
> 
> http://mail-archives.apache.org/mod_mbox/cloudstack-dev/201502.mbox/%3
c54eef6be.5040...@shapeblue.com%3E>
>
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
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=Hhwq
-END PGP SIGNATURE-

[GitHub] cloudstack pull request: GetUsageRecordsCmd - NewInputFormat for S...

[svscorp]

GitHub user svscorp opened a pull request:

https://github.com/apache/cloudstack/pull/185

GetUsageRecordsCmd - NewInputFormat for START_DATE and END_DATE

It is extremely useful to be able to narrow the output to the specific time 
range within a date.

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/svscorp/cloudstack 
GetUsageRecords-NewInputFormat-StartDate-EndDate

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/cloudstack/pull/185.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #185


commit a0c6b4c54c6dc788a417455b5e649dc81038ac83
Author: Ilia Shakitko 
Date:   2015-04-21T13:43:25Z

GetUsageRecordsCmd - NewInputFormat for START_DATE and END_DATE (with time; 
backward capabile)




---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

CentOS 7 KVM Agent installation fails

[Wido den Hollander]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Hi,

I'm trying to install cloudstack-agent on CentOS but it fails with:

- --> Finished Dependency Resolution
Error: Package: cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64
(cloudstack-repo)
   Requires: java7
Error: Package: cloudstack-common-4.4.2-NONOSS_1.el6.x86_64
(cloudstack-repo)
   Requires: python(abi) = 2.6
   Installed: python-2.7.5-16.el7.x86_64 (@anaconda)
   python(abi) = 2.7
   python(abi) = 2.7
 You could try using --skip-broken to work around the problem
 You could try running: rpm -Va --nofiles --nodigest


Looking at packaging/centos7/cloud.spec I found this:

%package agent
Summary: CloudStack Agent for KVM hypervisors
.
Requires: java => 1.7.0
.

So where does "java7" come from?

I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/

Any clues on why this is happening?

Wido
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
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=dRkq
-END PGP SIGNATURE-

Re: CentOS 7 KVM Agent installation fails

[Remi Bergsma]

Hi Wido,

We run kvm on centos7 without problems. When I'm back in the office I will 
check which package we used. It seems el6 is for centos6?

Will let you know!

Remi

Sent from my iPhone

> On 21 Apr 2015, at 16:00, Wido den Hollander  wrote:
> 
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Hi,
> 
> I'm trying to install cloudstack-agent on CentOS but it fails with:
> 
> - --> Finished Dependency Resolution
> Error: Package: cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64
> (cloudstack-repo)
>   Requires: java7
> Error: Package: cloudstack-common-4.4.2-NONOSS_1.el6.x86_64
> (cloudstack-repo)
>   Requires: python(abi) = 2.6
>   Installed: python-2.7.5-16.el7.x86_64 (@anaconda)
>   python(abi) = 2.7
>   python(abi) = 2.7
> You could try using --skip-broken to work around the problem
> You could try running: rpm -Va --nofiles --nodigest
> 
> 
> Looking at packaging/centos7/cloud.spec I found this:
> 
> %package agent
> Summary: CloudStack Agent for KVM hypervisors
> .
> Requires: java => 1.7.0
> .
> 
> So where does "java7" come from?
> 
> I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/
> 
> Any clues on why this is happening?
> 
> Wido
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1
> 
> iQIcBAEBAgAGBQJVNlgXAAoJEAGbWC3bPspCxToP/jyPrXpHgwW48A/WvbLZPCnw
> m1Q6YrVi0oP6hGxZjFkESSIvsyg60zo6Cx440XoCAmJTREVeLtbtTiQUL42Z5EcH
> Ko/wIzHJp/eoPtuPBk08m10pC6OgQitQFVO4xTieQxGWtcy+mBi+5VfWHszR0XEa
> PscABN82pveE8igf8TTOZtHkkXXdzILY5tkiq0j96ZN8cGNTVV0amcyraE0rxSFn
> n80ou84e98ek32QyC4wo9xBcRjl5aKu5/5oUeLESxJQWK6KEcjW0JR+4tOUZbLwp
> LMZJXnKNEpf+Lv/fHFAjz9RjjD1T96vERbyBIJGRqsgwKKNQ4XdJph4oh2GDKXYn
> cLKpfAFVGWn0Po+2m5PQDFgdi7rgHAvPieIFmaBbx90i3s6vV2SHVzNEhh4EKEL8
> hLPRZjA18wZ5wChnY2Lxug7OjqyBSvxuj5hcox8VntDR8wufcAlaBWMrLcZ0YTw0
> dP+7VyMckSHuPyeGragwF4oKw5ccXJWERg1ZdNpT3RX7axBbv8J+5HynbJ9zCE7n
> eW5zQlrVFZPLVUUolhmWzl12CfJspYbmQdAUlOon9mzLIb2SDR0eRr5xcvjEdUNK
> UZvtR1B60nbEcrEckdSEsuKBew9yxO4hzT0aD1BgFWtChNS04/hAnu/4smpry4bC
> QknkSCN6G6zoibyXYCjg
> =dRkq
> -END PGP SIGNATURE-

Re: CentOS 7 KVM Agent installation fails

[Wido den Hollander]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On 04/21/2015 04:10 PM, Remi Bergsma wrote:
> Hi Wido,
> 
> We run kvm on centos7 without problems. When I'm back in the office
> I will check which package we used. It seems el6 is for centos6?
> 

Thanks! Aha, let me check if the packages coming from Jenkins do work.

Wido

> Will let you know!
> 
> Remi
> 
> Sent from my iPhone
> 
>> On 21 Apr 2015, at 16:00, Wido den Hollander 
>> wrote:
>> 
> Hi,
> 
> I'm trying to install cloudstack-agent on CentOS but it fails
> with:
> 
> --> Finished Dependency Resolution Error: Package:
> cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo) 
> Requires: java7 Error: Package:
> cloudstack-common-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo) 
> Requires: python(abi) = 2.6 Installed: python-2.7.5-16.el7.x86_64
> (@anaconda) python(abi) = 2.7 python(abi) = 2.7 You could try using
> --skip-broken to work around the problem You could try running: rpm
> -Va --nofiles --nodigest
> 
> 
> Looking at packaging/centos7/cloud.spec I found this:
> 
> %package agent Summary: CloudStack Agent for KVM hypervisors . 
> Requires: java => 1.7.0 .
> 
> So where does "java7" come from?
> 
> I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/
> 
> Any clues on why this is happening?
> 
> Wido
-BEGIN PGP SIGNATURE-
Version: GnuPG v1
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=ogdp
-END PGP SIGNATURE-

SG not working with 6.5 and PR review

[Rohit Yadav]

Hi all,

We discovered that Security Groups don’t work in ACS 4.5+ when used with 
XenServer 6.5 due to ipset, sm/util.py changes. I’ve opened the issue here 
which was found to be reproducible by my colleagues Geoff and Abhi:
https://issues.apache.org/jira/browse/CLOUDSTACK-8395

I’ve tried to fix it in a way such that vmops plugin would work on both XS 6.2 
and 6.5 releases, here's the PR: https://github.com/apache/cloudstack/pull/186

One of the major changes it introduces it to use “nethash” instead of “iphash” 
when storing CIDRs received as part of a ingress/egress rule. I’m not sure how 
it will affect users that will upgrade to ACS 4.5, as a precaution I’ve added a 
change to flush and remove old ipset entry before adding a new one. (Assuming 
all network rule addition/removals are idempotent, as everytime we add/remove a 
rule, all rules are sent to be applied by the XS vmops plugins).

Tim - since you’re one of the Xen gurus can you help review it and suggest any 
other changes?

I wanted to bring this issue on dev ML since it’s a potential blocker for 4.5. 
I’m not sure if we officially support XS 6.5 on 4.4 branch, but if needed once 
we have a reviewed commit it can be cherry-picked on 4.4 as well.

Regards,
Rohit Yadav
Software Architect, ShapeBlue
M. +91 88 262 30892 | rohit.ya...@shapeblue.com
Blog: bhaisaab.org | Twitter: @_bhaisaab



Find out more about ShapeBlue and our range of CloudStack related services

IaaS Cloud Design & Build
CSForge – rapid IaaS deployment framework
CloudStack Consulting
CloudStack Software 
Engineering
CloudStack Infrastructure 
Support
CloudStack Bootcamp Training Courses

This email and any attachments to it may be confidential and are intended 
solely for the use of the individual to whom it is addressed. Any views or 
opinions expressed are solely those of the author and do not necessarily 
represent those of Shape Blue Ltd or related companies. If you are not the 
intended recipient of this email, you must neither take any action based upon 
its contents, nor copy or show it to anyone. Please contact the sender if you 
believe you have received this email in error. Shape Blue Ltd is a company 
incorporated in England & Wales. ShapeBlue Services India LLP is a company 
incorporated in India and is operated under license from Shape Blue Ltd. Shape 
Blue Brasil Consultoria Ltda is a company incorporated in Brasil and is 
operated under license from Shape Blue Ltd. ShapeBlue SA Pty Ltd is a company 
registered by The Republic of South Africa and is traded under license from 
Shape Blue Ltd. ShapeBlue is a registered trademark.

[GitHub] cloudstack pull request: CLOUDSTACK-8395: vmops plugin should work...

[bhaisaab]

GitHub user bhaisaab opened a pull request:

https://github.com/apache/cloudstack/pull/186

CLOUDSTACK-8395: vmops plugin should work on both XS 6.5 and 6.2 :fist:

This fixes the issue of Security Groups not working in case of XenServer 
6.5;
- Uses nethash ipset data-structure to store CIDRs (efficient than iphash 
and
  avoids overflow errors in case users add /8 /4 ingress/egress cidrs)
- Support for ipset versions both on 6.2 and 6.5, both have different 
outputs. This
  fixes the issue of destroy_network_rules_for_vm failing
- Implements defensive filtering of list, instead of popping last item 
without
  checking if it's None or empty
- Greps using names that are 'quoted' to avoid bash errors
- Before setting up new network rule, tries to clean and remove old ipset 
entry
- Idents, whitespace and naming fixes

PS. This is my 1000th commit to the :monkey_face: project :)

Signed-off-by: Rohit Yadav 

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/apache/cloudstack CLOUDSTACK-8395

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/cloudstack/pull/186.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #186


commit c11080a99d0479c51eaa4f2d2664ed7f8e348a5a
Author: Rohit Yadav 
Date:   2015-04-21T15:35:36Z

CLOUDSTACK-8395: vmops plugin should work on both XS 6.5 and 6.2 :fist:

This fixes the issue of Security Groups not working in case of XenServer 
6.5;
- Uses nethash ipset data-structure to store CIDRs (efficient than iphash 
and
  avoids overflow errors in case users add /8 /4 ingress/egress cidrs)
- Support for ipset versions both on 6.2 and 6.5, both have different 
outputs. This
  fixes the issue of destroy_network_rules_for_vm failing
- Implements defensive filtering of list, instead of popping last item 
without
  checking if it's None or empty
- Greps using names that are 'quoted' to avoid bash errors
- Before setting up new network rule, tries to clean and remove old ipset 
entry
- Idents, whitespace and naming fixes

PS. This is my 1000th commit to the :monkey_face: project :)

Signed-off-by: Rohit Yadav 




---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

Re: CentOS 7 KVM Agent installation fails

[Rohit Yadav]

Hi Wido,

You can try the latest 4.4.3 release from this EL6 repo: 
http://packages.shapeblue.com/cloudstack/upstream/centos/4.4

> On 21-Apr-2015, at 4:25 pm, Wido den Hollander  wrote:
>
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> On 04/21/2015 04:10 PM, Remi Bergsma wrote:
>> Hi Wido,
>>
>> We run kvm on centos7 without problems. When I'm back in the office
>> I will check which package we used. It seems el6 is for centos6?
>>
>
> Thanks! Aha, let me check if the packages coming from Jenkins do work.
>
> Wido
>
>> Will let you know!
>>
>> Remi
>>
>> Sent from my iPhone
>>
>>> On 21 Apr 2015, at 16:00, Wido den Hollander 
>>> wrote:
>>>
>> Hi,
>>
>> I'm trying to install cloudstack-agent on CentOS but it fails
>> with:
>>
>> --> Finished Dependency Resolution Error: Package:
>> cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
>> Requires: java7 Error: Package:
>> cloudstack-common-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
>> Requires: python(abi) = 2.6 Installed: python-2.7.5-16.el7.x86_64
>> (@anaconda) python(abi) = 2.7 python(abi) = 2.7 You could try using
>> --skip-broken to work around the problem You could try running: rpm
>> -Va --nofiles --nodigest
>>
>>
>> Looking at packaging/centos7/cloud.spec I found this:
>>
>> %package agent Summary: CloudStack Agent for KVM hypervisors .
>> Requires: java => 1.7.0 .
>>
>> So where does "java7" come from?
>>
>> I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/
>>
>> Any clues on why this is happening?
>>
>> Wido
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1
>
> iQIcBAEBAgAGBQJVNl3kAAoJEAGbWC3bPspCbSUQAI1E/Rk7nByBiaXM0oTMPwWK
> zefmTQ2R5at6RBqR1zM8HSR8pNiyUn53uua0/FIhKy9HmoRhNjw5BDIDzpz4RcgY
> HNUSRyaAI+YqI5g4FcDalsc+Vqelcv9zEzvksG5VkspESrdhJTEX73FLWiBUr7oM
> FemnBO8smX7Pc2awZRS/XBjoGq8GISmUO3mIHmT2UFyjhynlRGehhOhQGyLWjN9a
> gxLs7vtkKFdBuinsWny9NsX6D4RSnMoWAZ0h4fdDYINbzh24ZBBNiRhCPGaya0We
> 0YY8/WBy8CIQ5IH42/DoPbRdlZL/5744MdTaVU/ZkkQ6J4T9T6UFECk19YptwVgF
> Aq9E3nsuxwDeW+8TksHscy/0/Z8HKQGSa9YqqSneAYQGKNdjXylroc7/JcH7mSpo
> /RvS60l2k4NQ9mJrp8ta8rrwzahqiBLUKBQEs9KA0OqR67uV+p6SSXaVWEqx+819
> bjFqZR1rm2GBoS6xFITZpq1G9eB3RNe7ak7Nzdv5LfgQ4RMbtHvQ+DaSL8bAnyQo
> EMDbe4J7IQGLIQtfmeUnZsz2Lmz02XdYtzIFBc9CKP3EGYJSKDVqAQH4dGwPBkBY
> 9c56ggRXgLBtKcDQijZNM/h6geP6C2fAP/HdLjx0qfydR5ktML66dW6IFN3aAgwW
> lFlQurU8D0Ug0uW/T/0k
> =ogdp
> -END PGP SIGNATURE-

Regards,
Rohit Yadav
Software Architect, ShapeBlue
M. +91 88 262 30892 | rohit.ya...@shapeblue.com
Blog: bhaisaab.org | Twitter: @_bhaisaab



Find out more about ShapeBlue and our range of CloudStack related services

IaaS Cloud Design & Build
CSForge – rapid IaaS deployment framework
CloudStack Consulting
CloudStack Software 
Engineering
CloudStack Infrastructure 
Support
CloudStack Bootcamp Training Courses

This email and any attachments to it may be confidential and are intended 
solely for the use of the individual to whom it is addressed. Any views or 
opinions expressed are solely those of the author and do not necessarily 
represent those of Shape Blue Ltd or related companies. If you are not the 
intended recipient of this email, you must neither take any action based upon 
its contents, nor copy or show it to anyone. Please contact the sender if you 
believe you have received this email in error. Shape Blue Ltd is a company 
incorporated in England & Wales. ShapeBlue Services India LLP is a company 
incorporated in India and is operated under license from Shape Blue Ltd. Shape 
Blue Brasil Consultoria Ltda is a company incorporated in Brasil and is 
operated under license from Shape Blue Ltd. ShapeBlue SA Pty Ltd is a company 
registered by The Republic of South Africa and is traded under license from 
Shape Blue Ltd. ShapeBlue is a registered trademark.

Re: CentOS 7 KVM Agent installation fails

[Remi Bergsma]

There are some fixes in 4.4.3 for kvm on CentOS7 so that's a good idea.

Regards, Remi 

Sent from my iPhone

> On 21 Apr 2015, at 18:09, Rohit Yadav  wrote:
> 
> Hi Wido,
> 
> You can try the latest 4.4.3 release from this EL6 repo: 
> http://packages.shapeblue.com/cloudstack/upstream/centos/4.4
> 
>> On 21-Apr-2015, at 4:25 pm, Wido den Hollander  wrote:
>> 
>> -BEGIN PGP SIGNED MESSAGE-
>> Hash: SHA1
>> 
>>> On 04/21/2015 04:10 PM, Remi Bergsma wrote:
>>> Hi Wido,
>>> 
>>> We run kvm on centos7 without problems. When I'm back in the office
>>> I will check which package we used. It seems el6 is for centos6?
>> 
>> Thanks! Aha, let me check if the packages coming from Jenkins do work.
>> 
>> Wido
>> 
>>> Will let you know!
>>> 
>>> Remi
>>> 
>>> Sent from my iPhone
>>> 
 On 21 Apr 2015, at 16:00, Wido den Hollander 
 wrote:
>>> Hi,
>>> 
>>> I'm trying to install cloudstack-agent on CentOS but it fails
>>> with:
>>> 
>>> --> Finished Dependency Resolution Error: Package:
>>> cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
>>> Requires: java7 Error: Package:
>>> cloudstack-common-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
>>> Requires: python(abi) = 2.6 Installed: python-2.7.5-16.el7.x86_64
>>> (@anaconda) python(abi) = 2.7 python(abi) = 2.7 You could try using
>>> --skip-broken to work around the problem You could try running: rpm
>>> -Va --nofiles --nodigest
>>> 
>>> 
>>> Looking at packaging/centos7/cloud.spec I found this:
>>> 
>>> %package agent Summary: CloudStack Agent for KVM hypervisors .
>>> Requires: java => 1.7.0 .
>>> 
>>> So where does "java7" come from?
>>> 
>>> I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/
>>> 
>>> Any clues on why this is happening?
>>> 
>>> Wido
>> -BEGIN PGP SIGNATURE-
>> Version: GnuPG v1
>> 
>> iQIcBAEBAgAGBQJVNl3kAAoJEAGbWC3bPspCbSUQAI1E/Rk7nByBiaXM0oTMPwWK
>> zefmTQ2R5at6RBqR1zM8HSR8pNiyUn53uua0/FIhKy9HmoRhNjw5BDIDzpz4RcgY
>> HNUSRyaAI+YqI5g4FcDalsc+Vqelcv9zEzvksG5VkspESrdhJTEX73FLWiBUr7oM
>> FemnBO8smX7Pc2awZRS/XBjoGq8GISmUO3mIHmT2UFyjhynlRGehhOhQGyLWjN9a
>> gxLs7vtkKFdBuinsWny9NsX6D4RSnMoWAZ0h4fdDYINbzh24ZBBNiRhCPGaya0We
>> 0YY8/WBy8CIQ5IH42/DoPbRdlZL/5744MdTaVU/ZkkQ6J4T9T6UFECk19YptwVgF
>> Aq9E3nsuxwDeW+8TksHscy/0/Z8HKQGSa9YqqSneAYQGKNdjXylroc7/JcH7mSpo
>> /RvS60l2k4NQ9mJrp8ta8rrwzahqiBLUKBQEs9KA0OqR67uV+p6SSXaVWEqx+819
>> bjFqZR1rm2GBoS6xFITZpq1G9eB3RNe7ak7Nzdv5LfgQ4RMbtHvQ+DaSL8bAnyQo
>> EMDbe4J7IQGLIQtfmeUnZsz2Lmz02XdYtzIFBc9CKP3EGYJSKDVqAQH4dGwPBkBY
>> 9c56ggRXgLBtKcDQijZNM/h6geP6C2fAP/HdLjx0qfydR5ktML66dW6IFN3aAgwW
>> lFlQurU8D0Ug0uW/T/0k
>> =ogdp
>> -END PGP SIGNATURE-
> 
> Regards,
> Rohit Yadav
> Software Architect, ShapeBlue
> M. +91 88 262 30892 | rohit.ya...@shapeblue.com
> Blog: bhaisaab.org | Twitter: @_bhaisaab
> 
> 
> 
> Find out more about ShapeBlue and our range of CloudStack related services
> 
> IaaS Cloud Design & Build
> CSForge – rapid IaaS deployment framework
> CloudStack Consulting
> CloudStack Software 
> Engineering
> CloudStack Infrastructure 
> Support
> CloudStack Bootcamp Training 
> Courses
> 
> This email and any attachments to it may be confidential and are intended 
> solely for the use of the individual to whom it is addressed. Any views or 
> opinions expressed are solely those of the author and do not necessarily 
> represent those of Shape Blue Ltd or related companies. If you are not the 
> intended recipient of this email, you must neither take any action based upon 
> its contents, nor copy or show it to anyone. Please contact the sender if you 
> believe you have received this email in error. Shape Blue Ltd is a company 
> incorporated in England & Wales. ShapeBlue Services India LLP is a company 
> incorporated in India and is operated under license from Shape Blue Ltd. 
> Shape Blue Brasil Consultoria Ltda is a company incorporated in Brasil and is 
> operated under license from Shape Blue Ltd. ShapeBlue SA Pty Ltd is a company 
> registered by The Republic of South Africa and is traded under license from 
> Shape Blue Ltd. ShapeBlue is a registered trademark.

Re: CentOS 7 KVM Agent installation fails

[Marcus]

This requires has actually bitten us in the past. It needs to change
to 'java-1.7.0', because 'java7' isn't satisfied by anything in EL7,
and 'java >= 1.7.0' actually is satisfied by java 1.6 due to epoch
issues.

That said, EL7 support was considered 'tech preview' even with 4.5,
I'm not sure what the status of it is in 4.4.

Also, in 4.5 we use packaging/centos63/package.sh -o rhel7. I haven't
researched why/when this was added here or if it is applicable to 4.4,
but it works.

On Tue, Apr 21, 2015 at 9:17 AM, Remi Bergsma  wrote:
> There are some fixes in 4.4.3 for kvm on CentOS7 so that's a good idea.
>
> Regards, Remi
>
> Sent from my iPhone
>
>> On 21 Apr 2015, at 18:09, Rohit Yadav  wrote:
>>
>> Hi Wido,
>>
>> You can try the latest 4.4.3 release from this EL6 repo: 
>> http://packages.shapeblue.com/cloudstack/upstream/centos/4.4
>>
>>> On 21-Apr-2015, at 4:25 pm, Wido den Hollander  wrote:
>>>
>>> -BEGIN PGP SIGNED MESSAGE-
>>> Hash: SHA1
>>>
 On 04/21/2015 04:10 PM, Remi Bergsma wrote:
 Hi Wido,

 We run kvm on centos7 without problems. When I'm back in the office
 I will check which package we used. It seems el6 is for centos6?
>>>
>>> Thanks! Aha, let me check if the packages coming from Jenkins do work.
>>>
>>> Wido
>>>
 Will let you know!

 Remi

 Sent from my iPhone

> On 21 Apr 2015, at 16:00, Wido den Hollander 
> wrote:
 Hi,

 I'm trying to install cloudstack-agent on CentOS but it fails
 with:

 --> Finished Dependency Resolution Error: Package:
 cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
 Requires: java7 Error: Package:
 cloudstack-common-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
 Requires: python(abi) = 2.6 Installed: python-2.7.5-16.el7.x86_64
 (@anaconda) python(abi) = 2.7 python(abi) = 2.7 You could try using
 --skip-broken to work around the problem You could try running: rpm
 -Va --nofiles --nodigest


 Looking at packaging/centos7/cloud.spec I found this:

 %package agent Summary: CloudStack Agent for KVM hypervisors .
 Requires: java => 1.7.0 .

 So where does "java7" come from?

 I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/

 Any clues on why this is happening?

 Wido
>>> -BEGIN PGP SIGNATURE-
>>> Version: GnuPG v1
>>>
>>> iQIcBAEBAgAGBQJVNl3kAAoJEAGbWC3bPspCbSUQAI1E/Rk7nByBiaXM0oTMPwWK
>>> zefmTQ2R5at6RBqR1zM8HSR8pNiyUn53uua0/FIhKy9HmoRhNjw5BDIDzpz4RcgY
>>> HNUSRyaAI+YqI5g4FcDalsc+Vqelcv9zEzvksG5VkspESrdhJTEX73FLWiBUr7oM
>>> FemnBO8smX7Pc2awZRS/XBjoGq8GISmUO3mIHmT2UFyjhynlRGehhOhQGyLWjN9a
>>> gxLs7vtkKFdBuinsWny9NsX6D4RSnMoWAZ0h4fdDYINbzh24ZBBNiRhCPGaya0We
>>> 0YY8/WBy8CIQ5IH42/DoPbRdlZL/5744MdTaVU/ZkkQ6J4T9T6UFECk19YptwVgF
>>> Aq9E3nsuxwDeW+8TksHscy/0/Z8HKQGSa9YqqSneAYQGKNdjXylroc7/JcH7mSpo
>>> /RvS60l2k4NQ9mJrp8ta8rrwzahqiBLUKBQEs9KA0OqR67uV+p6SSXaVWEqx+819
>>> bjFqZR1rm2GBoS6xFITZpq1G9eB3RNe7ak7Nzdv5LfgQ4RMbtHvQ+DaSL8bAnyQo
>>> EMDbe4J7IQGLIQtfmeUnZsz2Lmz02XdYtzIFBc9CKP3EGYJSKDVqAQH4dGwPBkBY
>>> 9c56ggRXgLBtKcDQijZNM/h6geP6C2fAP/HdLjx0qfydR5ktML66dW6IFN3aAgwW
>>> lFlQurU8D0Ug0uW/T/0k
>>> =ogdp
>>> -END PGP SIGNATURE-
>>
>> Regards,
>> Rohit Yadav
>> Software Architect, ShapeBlue
>> M. +91 88 262 30892 | rohit.ya...@shapeblue.com
>> Blog: bhaisaab.org | Twitter: @_bhaisaab
>>
>>
>>
>> Find out more about ShapeBlue and our range of CloudStack related services
>>
>> IaaS Cloud Design & Build
>> CSForge – rapid IaaS deployment framework
>> CloudStack Consulting
>> CloudStack Software 
>> Engineering
>> CloudStack Infrastructure 
>> Support
>> CloudStack Bootcamp Training 
>> Courses
>>
>> This email and any attachments to it may be confidential and are intended 
>> solely for the use of the individual to whom it is addressed. Any views or 
>> opinions expressed are solely those of the author and do not necessarily 
>> represent those of Shape Blue Ltd or related companies. If you are not the 
>> intended recipient of this email, you must neither take any action based 
>> upon its contents, nor copy or show it to anyone. Please contact the sender 
>> if you believe you have received this email in error. Shape Blue Ltd is a 
>> company incorporated in England & Wales. ShapeBlue Services India LLP is a 
>> company incorporated in India and is operated under license from Shape Blue 
>> Ltd. Shape Blue Brasil Consultoria Ltda is a company incorporated in Brasil 
>> and is operated under license from Shape Blue Ltd. ShapeBlue SA Pty Ltd is a 
>> company registered by The Republic of South Africa and is traded under 
>> license from Shape Blue Ltd. ShapeBlue is a registered trad

Jenkins build is still unstable: simulator-4.5-singlerun #213

[jenkins]

See 

[GitHub] cloudstack pull request: CLOUDSTACK-8395: vmops plugin should work...

[xenserverarmy]

Github user xenserverarmy commented on the pull request:

https://github.com/apache/cloudstack/pull/186#issuecomment-94880329
  
This looks solid, but I'd update line 457 to tick-quote the vmchain as 
you've done elsewhere


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

Re: SG not working with 6.5 and PR review

[Tim Mackey]

Rohit,

I just added a comment to update line 457 to tick-quote the vmchain as
you've done elsewhere.  My main concern would be flushing the ipset while
the iptable entry still exists.

I am curious what in sm/util.py concerned you.  That's all storage
management code and should have nothing to do with security groups.  I also
diffed a 6.5 and 6.2 version which didn't show anything obvious to explain
a security group issue.

ipset definitely did change going from 4.5 to 6.11 to match our kernel
update.

-tim

On Tue, Apr 21, 2015 at 11:57 AM, Rohit Yadav 
wrote:

> Hi all,
>
> We discovered that Security Groups don’t work in ACS 4.5+ when used with
> XenServer 6.5 due to ipset, sm/util.py changes. I’ve opened the issue here
> which was found to be reproducible by my colleagues Geoff and Abhi:
> https://issues.apache.org/jira/browse/CLOUDSTACK-8395
>
> I’ve tried to fix it in a way such that vmops plugin would work on both XS
> 6.2 and 6.5 releases, here's the PR:
> https://github.com/apache/cloudstack/pull/186
>
> One of the major changes it introduces it to use “nethash” instead of
> “iphash” when storing CIDRs received as part of a ingress/egress rule. I’m
> not sure how it will affect users that will upgrade to ACS 4.5, as a
> precaution I’ve added a change to flush and remove old ipset entry before
> adding a new one. (Assuming all network rule addition/removals are
> idempotent, as everytime we add/remove a rule, all rules are sent to be
> applied by the XS vmops plugins).
>
> Tim - since you’re one of the Xen gurus can you help review it and suggest
> any other changes?
>
> I wanted to bring this issue on dev ML since it’s a potential blocker for
> 4.5. I’m not sure if we officially support XS 6.5 on 4.4 branch, but if
> needed once we have a reviewed commit it can be cherry-picked on 4.4 as
> well.
>
> Regards,
> Rohit Yadav
> Software Architect, ShapeBlue
> M. +91 88 262 30892 | rohit.ya...@shapeblue.com
> Blog: bhaisaab.org | Twitter: @_bhaisaab
>
>
>
> Find out more about ShapeBlue and our range of CloudStack related services
>
> IaaS Cloud Design & Build<
> http://shapeblue.com/iaas-cloud-design-and-build//>
> CSForge – rapid IaaS deployment framework
> CloudStack Consulting
> CloudStack Software Engineering<
> http://shapeblue.com/cloudstack-software-engineering/>
> CloudStack Infrastructure Support<
> http://shapeblue.com/cloudstack-infrastructure-support/>
> CloudStack Bootcamp Training Courses<
> http://shapeblue.com/cloudstack-training/>
>
> This email and any attachments to it may be confidential and are intended
> solely for the use of the individual to whom it is addressed. Any views or
> opinions expressed are solely those of the author and do not necessarily
> represent those of Shape Blue Ltd or related companies. If you are not the
> intended recipient of this email, you must neither take any action based
> upon its contents, nor copy or show it to anyone. Please contact the sender
> if you believe you have received this email in error. Shape Blue Ltd is a
> company incorporated in England & Wales. ShapeBlue Services India LLP is a
> company incorporated in India and is operated under license from Shape Blue
> Ltd. Shape Blue Brasil Consultoria Ltda is a company incorporated in Brasil
> and is operated under license from Shape Blue Ltd. ShapeBlue SA Pty Ltd is
> a company registered by The Republic of South Africa and is traded under
> license from Shape Blue Ltd. ShapeBlue is a registered trademark.
>

Jenkins build is still unstable: simulator-singlerun #1134

[jenkins]

See 

Re: CentOS 7 KVM Agent installation fails

[Remi Bergsma]

Hi,

The RPM’s I use were baked from the 4.4 branch (4.4.3-SNAPSHOT) a few weeks 
back. Hence, the now released 4.4.3 should also work. Marcus is right, the 
dependency is now on ‘java >= 1.7.0’.

CentOS 7 support was needed in our setup and Daan helped me backport the needed 
changes from to 4.4. This allowed me to test in a production-like environment 
and fix whatever did not work properly. These patches are included in 4.4.3. 
Even though it runs fine now, we will probably find some new issues over time 
so I can understand the ’tech preview’ status. 

If you need more info or help, drop me a line.

Regards,
Remi


> On 21 Apr 2015, at 18:38 , Marcus  wrote:
> 
> This requires has actually bitten us in the past. It needs to change
> to 'java-1.7.0', because 'java7' isn't satisfied by anything in EL7,
> and 'java >= 1.7.0' actually is satisfied by java 1.6 due to epoch
> issues.
> 
> That said, EL7 support was considered 'tech preview' even with 4.5,
> I'm not sure what the status of it is in 4.4.
> 
> Also, in 4.5 we use packaging/centos63/package.sh -o rhel7. I haven't
> researched why/when this was added here or if it is applicable to 4.4,
> but it works.
> 
> On Tue, Apr 21, 2015 at 9:17 AM, Remi Bergsma  wrote:
>> There are some fixes in 4.4.3 for kvm on CentOS7 so that's a good idea.
>> 
>> Regards, Remi
>> 
>> Sent from my iPhone
>> 
>>> On 21 Apr 2015, at 18:09, Rohit Yadav  wrote:
>>> 
>>> Hi Wido,
>>> 
>>> You can try the latest 4.4.3 release from this EL6 repo: 
>>> http://packages.shapeblue.com/cloudstack/upstream/centos/4.4
>>> 
 On 21-Apr-2015, at 4:25 pm, Wido den Hollander  wrote:
 
 -BEGIN PGP SIGNED MESSAGE-
 Hash: SHA1
 
> On 04/21/2015 04:10 PM, Remi Bergsma wrote:
> Hi Wido,
> 
> We run kvm on centos7 without problems. When I'm back in the office
> I will check which package we used. It seems el6 is for centos6?
 
 Thanks! Aha, let me check if the packages coming from Jenkins do work.
 
 Wido
 
> Will let you know!
> 
> Remi
> 
> Sent from my iPhone
> 
>> On 21 Apr 2015, at 16:00, Wido den Hollander 
>> wrote:
> Hi,
> 
> I'm trying to install cloudstack-agent on CentOS but it fails
> with:
> 
> --> Finished Dependency Resolution Error: Package:
> cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
> Requires: java7 Error: Package:
> cloudstack-common-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
> Requires: python(abi) = 2.6 Installed: python-2.7.5-16.el7.x86_64
> (@anaconda) python(abi) = 2.7 python(abi) = 2.7 You could try using
> --skip-broken to work around the problem You could try running: rpm
> -Va --nofiles --nodigest
> 
> 
> Looking at packaging/centos7/cloud.spec I found this:
> 
> %package agent Summary: CloudStack Agent for KVM hypervisors .
> Requires: java => 1.7.0 .
> 
> So where does "java7" come from?
> 
> I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/
> 
> Any clues on why this is happening?
> 
> Wido
 -BEGIN PGP SIGNATURE-
 Version: GnuPG v1
 
 iQIcBAEBAgAGBQJVNl3kAAoJEAGbWC3bPspCbSUQAI1E/Rk7nByBiaXM0oTMPwWK
 zefmTQ2R5at6RBqR1zM8HSR8pNiyUn53uua0/FIhKy9HmoRhNjw5BDIDzpz4RcgY
 HNUSRyaAI+YqI5g4FcDalsc+Vqelcv9zEzvksG5VkspESrdhJTEX73FLWiBUr7oM
 FemnBO8smX7Pc2awZRS/XBjoGq8GISmUO3mIHmT2UFyjhynlRGehhOhQGyLWjN9a
 gxLs7vtkKFdBuinsWny9NsX6D4RSnMoWAZ0h4fdDYINbzh24ZBBNiRhCPGaya0We
 0YY8/WBy8CIQ5IH42/DoPbRdlZL/5744MdTaVU/ZkkQ6J4T9T6UFECk19YptwVgF
 Aq9E3nsuxwDeW+8TksHscy/0/Z8HKQGSa9YqqSneAYQGKNdjXylroc7/JcH7mSpo
 /RvS60l2k4NQ9mJrp8ta8rrwzahqiBLUKBQEs9KA0OqR67uV+p6SSXaVWEqx+819
 bjFqZR1rm2GBoS6xFITZpq1G9eB3RNe7ak7Nzdv5LfgQ4RMbtHvQ+DaSL8bAnyQo
 EMDbe4J7IQGLIQtfmeUnZsz2Lmz02XdYtzIFBc9CKP3EGYJSKDVqAQH4dGwPBkBY
 9c56ggRXgLBtKcDQijZNM/h6geP6C2fAP/HdLjx0qfydR5ktML66dW6IFN3aAgwW
 lFlQurU8D0Ug0uW/T/0k
 =ogdp
 -END PGP SIGNATURE-
>>> 
>>> Regards,
>>> Rohit Yadav
>>> Software Architect, ShapeBlue
>>> M. +91 88 262 30892 | rohit.ya...@shapeblue.com
>>> Blog: bhaisaab.org | Twitter: @_bhaisaab
>>> 
>>> 
>>> 
>>> Find out more about ShapeBlue and our range of CloudStack related services
>>> 
>>> IaaS Cloud Design & 
>>> Build
>>> CSForge – rapid IaaS deployment framework
>>> CloudStack Consulting
>>> CloudStack Software 
>>> Engineering
>>> CloudStack Infrastructure 
>>> Support
>>> CloudStack Bootcamp Training 
>>> Courses
>>> 
>>> This email and any attachments to it may be confidential and are intended 
>>> solely for the use of the individual to whom it is addressed. Any views or 
>>> opini

[GitHub] cloudstack pull request: fix typo in rsyslog logrotate script on V...

[remibergsma]

Github user remibergsma closed the pull request at:

https://github.com/apache/cloudstack/pull/182


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

Re: Next ACS release?

[ilya]

Andrei,

To best of my knowledge, both 4.4.x and 4.5.x are being worked on 
actively. As a community, we need to get better on QA of each release - 
this is something we are planning to cover this year with distributed QA 
model, this was not widely discussed yet but something we need to tackle.


4.5 rc2 got stalled and we need to restart. We had a 4 month release 
cycle, but we can really stick to it hard - as its community driven. May 
will have to revise it down to 6 months or so.


Regards
ilya

On 4/21/15 1:26 AM, Andrei Mikhailovsky wrote:

Hello guys,

Looking at the dev and user lists it is becoming less certain if version 4.5.x 
is ever coming out. It seems like a few months have passed since the not so 
fortunate release of 4.5.0 and I can't find a release schedule for the 4.5.1, 
which seems to have stopped at rc2 stage and haven't progressed further to a 
release stage.

Are we likely to see any progress with the 4.5.x branch or is the community 
switching towards the 4.6.x branch without releasing the 4.5.x?

I am a bit unclear as there are no release dates, schedules or dead lines that 
the community should work with. Possibly as a result of this, the ACS releases 
are not being released on time or fast enough.

Does it make sense to introduce release schedules for ACS that the dev 
community should stick to? Similar to what is being done in many other 
projects, like Ubuntu, etc. Or would this break the ACS project releases even 
more?

Andrei

[GitHub] cloudstack pull request: CLOUDSTACK-8395: vmops plugin should work...

[bhaisaab]

Github user bhaisaab commented on the pull request:

https://github.com/apache/cloudstack/pull/186#issuecomment-94911018
  
@xenserverarmy thanks Tim for a quick review, I've fixed the case you've 
mentioned.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

getting rid of 3rd party repositories (ceph, libvirt)

[Laszlo Hornyak]

Hi,

I have uploaded the libvirt dependency to maven central and you should
never get the maven failure if the libvirt.org server goes down, it will be
downloaded from central. I have sent a PR (#180) to remove the libvirt.org
repository from the build.

With ceph.org the situation:
- I have requested permission to publish like with libvirt, my request was
rejected since the ceph.com team holds ownership
- The only solution from here is to change the groupId of the dependency.
- org.apache.* is very likely no go, it is owned by apache, artifacts are
synchonrized to central through the apache repository, which is for apache
artifacts only.
- we can upload with a new groupId - suggestions are welcome

Regards,
Laszlo
-- 

EOF

Announcing Apache™ CloudStack™ v4.4.3

[Daan Hoogland]

Mature, easy-to-deploy Open Source Cloud computing software platform
boasts improved efficiency and performance.

The Apache CloudStack project announced the immediate availability of
Apache CloudStack v4.4.3, the latest version of the turnkey Open
Source cloud computing software platform used for creating private-,
public-, and hybrid cloud environments.

Apache CloudStack clouds enable billions of dollars' worth of business
transactions annually across their clouds, and its maturity and
stability has led it to has become the Open Source platform for many
service providers to set up on-demand, elastic public cloud computing
services, as well as enterprises and others to set up a private or
hybrid cloud for use by their own employees.

"While working hard on a new release we managed to backport many
stabilizing fixes to this maintenance release. Both for 4.5 and the
next version work is going hard and I am happy that we managed to
bring out this fix release for users of prior versions."
Daan Hoogland, member of the Apache CloudStack Project Management
Committee and release manager for version 4.4.

Lauded by Gartner Group, CloudStack includes an intuitive user
interface and rich APIs for managing the compute, networking,
software, and storage infrastructure resources.

CloudStack v4.4.3 reflects dozens of fixes

A complete overview of all new enhancements can be found in the
project release notes at
http://docs.cloudstack.apache.org/projects/cloudstack-release-notes/en/4.4.3/

CloudStack has been used by thousands of organizations worldwide and
is in use/production at Alcatel-Lucent, Autodesk, BT Cloud, China
Telecom, DATACAENTER Services, DataPipe, Edmunds.com, Exoscale,
GreenQloud, Hokkaido University, IDC Frontier, Ikoula, KDDI, KT/Korea
Telecom, LeaseWeb, NTT, Orange, PCextreme, Schuberg Philis, Shopzilla,
Slovak Telekom, SunGard AS, Taiwan Mobile, Tata, Trader Media Group,
TomTom, University of Melbourne, University of Sao Paolo, Verizon,
WebMD and Zynga, among others.

CloudStack originated at Cloud.com, which was acquired by Citrix in
2011. CloudStack was submitted to the Apache Incubator in April 2012
and graduated as an Apache Software Foundation Top-level Project in
March 2013.

Availability

CloudStack v4.4.3 is available immediately as a free download from
http://cloudstack.apache.org/downloads.html. Apache CloudStack
software is released under the Apache License v2.0.

Governance and Oversight

Apache CloudStack is overseen by a self-selected team of active
contributors to the project. A Project Management Committee (PMC)
guides the Project's day-to-day operations, including community
development and product releases.

Get Involved!

Apache CloudStack welcomes contribution and community participation
through mailing lists as well as attending face-to-face MeetUps,
developer trainings, and user events. Catch CloudStack in action at
the CloudStack Days

CloudStack Days Tokyo 2015, 2nd June 2015,
http://events.linuxfoundation.org/events/cloudstack-tokyo

CloudStack Days Seattle 2015 , 20th August 2015,
http://events.linuxfoundation.org/events/cloudstack-seattle

and

CloudStack Collaboration Conference 2015 in Dublin, 8th and 9th of
October 2015, http://cloudstackcollab.org

-- 
Daan

Re: Announcing Apache™ CloudStack™ v4.4.3

[Remi Bergsma]

Hi Daan,

Awesome!

Could you have a look at the release notes link? It’s a 404 now.

Regards,
Remi


> On 21 Apr 2015, at 21:29 , Daan Hoogland  wrote:
> 
> Mature, easy-to-deploy Open Source Cloud computing software platform
> boasts improved efficiency and performance.
> 
> The Apache CloudStack project announced the immediate availability of
> Apache CloudStack v4.4.3, the latest version of the turnkey Open
> Source cloud computing software platform used for creating private-,
> public-, and hybrid cloud environments.
> 
> Apache CloudStack clouds enable billions of dollars' worth of business
> transactions annually across their clouds, and its maturity and
> stability has led it to has become the Open Source platform for many
> service providers to set up on-demand, elastic public cloud computing
> services, as well as enterprises and others to set up a private or
> hybrid cloud for use by their own employees.
> 
> "While working hard on a new release we managed to backport many
> stabilizing fixes to this maintenance release. Both for 4.5 and the
> next version work is going hard and I am happy that we managed to
> bring out this fix release for users of prior versions."
> Daan Hoogland, member of the Apache CloudStack Project Management
> Committee and release manager for version 4.4.
> 
> Lauded by Gartner Group, CloudStack includes an intuitive user
> interface and rich APIs for managing the compute, networking,
> software, and storage infrastructure resources.
> 
> CloudStack v4.4.3 reflects dozens of fixes
> 
> A complete overview of all new enhancements can be found in the
> project release notes at
> http://docs.cloudstack.apache.org/projects/cloudstack-release-notes/en/4.4.3/
> 
> CloudStack has been used by thousands of organizations worldwide and
> is in use/production at Alcatel-Lucent, Autodesk, BT Cloud, China
> Telecom, DATACAENTER Services, DataPipe, Edmunds.com, Exoscale,
> GreenQloud, Hokkaido University, IDC Frontier, Ikoula, KDDI, KT/Korea
> Telecom, LeaseWeb, NTT, Orange, PCextreme, Schuberg Philis, Shopzilla,
> Slovak Telekom, SunGard AS, Taiwan Mobile, Tata, Trader Media Group,
> TomTom, University of Melbourne, University of Sao Paolo, Verizon,
> WebMD and Zynga, among others.
> 
> CloudStack originated at Cloud.com, which was acquired by Citrix in
> 2011. CloudStack was submitted to the Apache Incubator in April 2012
> and graduated as an Apache Software Foundation Top-level Project in
> March 2013.
> 
> Availability
> 
> CloudStack v4.4.3 is available immediately as a free download from
> http://cloudstack.apache.org/downloads.html. Apache CloudStack
> software is released under the Apache License v2.0.
> 
> Governance and Oversight
> 
> Apache CloudStack is overseen by a self-selected team of active
> contributors to the project. A Project Management Committee (PMC)
> guides the Project's day-to-day operations, including community
> development and product releases.
> 
> Get Involved!
> 
> Apache CloudStack welcomes contribution and community participation
> through mailing lists as well as attending face-to-face MeetUps,
> developer trainings, and user events. Catch CloudStack in action at
> the CloudStack Days
> 
> CloudStack Days Tokyo 2015, 2nd June 2015,
> http://events.linuxfoundation.org/events/cloudstack-tokyo
> 
> CloudStack Days Seattle 2015 , 20th August 2015,
> http://events.linuxfoundation.org/events/cloudstack-seattle
> 
> and
> 
> CloudStack Collaboration Conference 2015 in Dublin, 8th and 9th of
> October 2015, http://cloudstackcollab.org
> 
> -- 
> Daan

Re: SG not working with 6.5 and PR review

[Rohit Yadav]

Hi Tim,

Thanks for your review. I’ve fixed the tick-quote issue on the mentioned line. 
Please find the various issues addressed in this patch;

## 1. Issue with sm/util.py

My concern with sm/util.py was that, previously on XS 6.2 the output of a 
typical util.pread2().split(‘\n’) would return a list like 
[a,b,c,d,”"] with the last element as an empty string possibly because an extra 
newline in the command output. Due to this all the code in vmops plugin 
followed this kind of patten where the programmer coded it to do a list.pop() 
after calling pread2().split(‘\n’).

But in case of XS 6.5, I found that no newline was at the end and we should not 
assume it and doing a list.pop() would remove the last element which can be a 
valid string and not a newline (or empty string after calling split() on it). I 
traced this issue when I saw following methods fail in the 
/var/log/cloud/cloud.log and looked at the code:
delete_rules_for_vm_in_bridge_firewall_chain
destroy_ebtables_rules
destroy_arptables_rules

The fix I tried was to filter out empty and None elements from the list using a 
"filter(util.pread2().split(‘\n’))” kind of pattern instead of doing 
a list.pop() which might remove a valid string. After this fix, I saw that SG 
rules applied correctly and the previous errors went away. I believe that this 
way of defensive coding would work on both XS 6.2 and 6.5, irrespective of the 
output having empty lines or not.


## 2. Issue with large CIDRs and ipset overflow

The other issue I found was when I added an ingress or egress CIDR that was in 
the block /8 or /4. I found that ipset failed in the /var/log/cloud/cloud.log. 
On tracing the issue I found that ipset data structure has been modified to use 
iphash instead of iptreemap (done in bd6f03aa954d4b3e7ead7e8010c5674d5d1f9513) 
because recent versions of ipset (as in XS 6.5) did not have that 
data-structure.

The issue with “iphash” in this case was that when someone added a CIDR or 
CIDRs, it would add individual IPs instead of the CIDR itself. Which failed for 
/8 or /4 because the default max length of “iphash” entry was 65535. The fix 
for this was to use “nethash” since we receive CIDRs only and it becomes 
memory/space efficient. Only the ipset entry for the VM itself is “iphash” 
based now (to store a VM’s primary and secondary IPs).


## 3. Issue with removing Ingress/Egress rules

I saw that when an ingress or egress rule was removed while the iptable rules 
were removed/fixed, the ipset entry stayed there and cidrs in it were not 
getting removed. Every time I added or removed an ingress or egress entry, 
CloudStack would send the vmops plugin all the ingress and egress rules. This 
assumed that all add/remove operations were sort of idempotent and all 
information was applied every time we made any change. Therefore, as a 
pessimistic approach I added the fix to flush/remove the ipset entries before 
adding them back again (that the mgmt server sent to vmops plugin). This also 
solves an upgrade error for users who would upgrade to XS 6.5 or ACS 4.5, as 
previous ipset entries were of type “iptreemap” or “iphash”, but new entries 
(the ingress and egress ones) is of “nethash” type.


## Discuss - should we flush/remove ipset rules before applying new rules?

Can you advise if we should have this pessimistic approach (as it flushes ipset 
entries when rules may be still in place?) or use an optimistic approach to 
keep the entries. There are two issues with the current optimistic approach:
1. When egress/ingress rules are removed, for a brief period (until new ipset 
entry replace the old one) old entries still persist. But since iptables  rules 
are removed and no one is referencing them it is harmless.
2. In case of users upgrading, the old ipset data-structure is “iptreemap” or 
“iphash”, but the new one is “nethash”, how do we make sure that old ones are 
removed/deleted so new ones are of “nethash” type.

> On 21-Apr-2015, at 7:51 pm, Tim Mackey  wrote:
>
> Rohit,
>
> I just added a comment to update line 457 to tick-quote the vmchain as
> you've done elsewhere.  My main concern would be flushing the ipset while
> the iptable entry still exists.
>
> I am curious what in sm/util.py concerned you.  That's all storage
> management code and should have nothing to do with security groups.  I also
> diffed a 6.5 and 6.2 version which didn't show anything obvious to explain
> a security group issue.
>
> ipset definitely did change going from 4.5 to 6.11 to match our kernel
> update.
>
> -tim
>
> On Tue, Apr 21, 2015 at 11:57 AM, Rohit Yadav 
> wrote:
>
>> Hi all,
>>
>> We discovered that Security Groups don’t work in ACS 4.5+ when used with
>> XenServer 6.5 due to ipset, sm/util.py changes. I’ve opened the issue here
>> which was found to be reproducible by my colleagues Geoff and Abhi:
>> https://issues.apache.org/jira/browse/CLOUDSTACK-8395
>>
>> I’ve tried to fix it in a way such that vmops plugin would work on both XS
>> 6.2 and 6.5 

Re: CentOS 7 KVM Agent installation fails

[Remi Bergsma]

Hi Wido,

Take them from here:
http://jenkins.buildacloud.org/view/4.4/job/package-centos7-4.4-noredist/ 


or here:
http://cloudstack.apt-get.eu/centos/4.4/7/ 


Worked for me.

Regards,
Remi


> On 21 Apr 2015, at 20:10 , Remi Bergsma  wrote:
> 
> Hi,
> 
> The RPM’s I use were baked from the 4.4 branch (4.4.3-SNAPSHOT) a few weeks 
> back. Hence, the now released 4.4.3 should also work. Marcus is right, the 
> dependency is now on ‘java >= 1.7.0’.
> 
> CentOS 7 support was needed in our setup and Daan helped me backport the 
> needed changes from to 4.4. This allowed me to test in a production-like 
> environment and fix whatever did not work properly. These patches are 
> included in 4.4.3. Even though it runs fine now, we will probably find some 
> new issues over time so I can understand the ’tech preview’ status. 
> 
> If you need more info or help, drop me a line.
> 
> Regards,
> Remi
> 
> 
>> On 21 Apr 2015, at 18:38 , Marcus  wrote:
>> 
>> This requires has actually bitten us in the past. It needs to change
>> to 'java-1.7.0', because 'java7' isn't satisfied by anything in EL7,
>> and 'java >= 1.7.0' actually is satisfied by java 1.6 due to epoch
>> issues.
>> 
>> That said, EL7 support was considered 'tech preview' even with 4.5,
>> I'm not sure what the status of it is in 4.4.
>> 
>> Also, in 4.5 we use packaging/centos63/package.sh -o rhel7. I haven't
>> researched why/when this was added here or if it is applicable to 4.4,
>> but it works.
>> 
>> On Tue, Apr 21, 2015 at 9:17 AM, Remi Bergsma  wrote:
>>> There are some fixes in 4.4.3 for kvm on CentOS7 so that's a good idea.
>>> 
>>> Regards, Remi
>>> 
>>> Sent from my iPhone
>>> 
 On 21 Apr 2015, at 18:09, Rohit Yadav  wrote:
 
 Hi Wido,
 
 You can try the latest 4.4.3 release from this EL6 repo: 
 http://packages.shapeblue.com/cloudstack/upstream/centos/4.4
 
> On 21-Apr-2015, at 4:25 pm, Wido den Hollander  wrote:
> 
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
>> On 04/21/2015 04:10 PM, Remi Bergsma wrote:
>> Hi Wido,
>> 
>> We run kvm on centos7 without problems. When I'm back in the office
>> I will check which package we used. It seems el6 is for centos6?
> 
> Thanks! Aha, let me check if the packages coming from Jenkins do work.
> 
> Wido
> 
>> Will let you know!
>> 
>> Remi
>> 
>> Sent from my iPhone
>> 
>>> On 21 Apr 2015, at 16:00, Wido den Hollander 
>>> wrote:
>> Hi,
>> 
>> I'm trying to install cloudstack-agent on CentOS but it fails
>> with:
>> 
>> --> Finished Dependency Resolution Error: Package:
>> cloudstack-agent-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
>> Requires: java7 Error: Package:
>> cloudstack-common-4.4.2-NONOSS_1.el6.x86_64 (cloudstack-repo)
>> Requires: python(abi) = 2.6 Installed: python-2.7.5-16.el7.x86_64
>> (@anaconda) python(abi) = 2.7 python(abi) = 2.7 You could try using
>> --skip-broken to work around the problem You could try running: rpm
>> -Va --nofiles --nodigest
>> 
>> 
>> Looking at packaging/centos7/cloud.spec I found this:
>> 
>> %package agent Summary: CloudStack Agent for KVM hypervisors .
>> Requires: java => 1.7.0 .
>> 
>> So where does "java7" come from?
>> 
>> I used these packages btw: http://cloudstack.apt-get.eu/rhel/4.4/
>> 
>> Any clues on why this is happening?
>> 
>> Wido
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1
> 
> iQIcBAEBAgAGBQJVNl3kAAoJEAGbWC3bPspCbSUQAI1E/Rk7nByBiaXM0oTMPwWK
> zefmTQ2R5at6RBqR1zM8HSR8pNiyUn53uua0/FIhKy9HmoRhNjw5BDIDzpz4RcgY
> HNUSRyaAI+YqI5g4FcDalsc+Vqelcv9zEzvksG5VkspESrdhJTEX73FLWiBUr7oM
> FemnBO8smX7Pc2awZRS/XBjoGq8GISmUO3mIHmT2UFyjhynlRGehhOhQGyLWjN9a
> gxLs7vtkKFdBuinsWny9NsX6D4RSnMoWAZ0h4fdDYINbzh24ZBBNiRhCPGaya0We
> 0YY8/WBy8CIQ5IH42/DoPbRdlZL/5744MdTaVU/ZkkQ6J4T9T6UFECk19YptwVgF
> Aq9E3nsuxwDeW+8TksHscy/0/Z8HKQGSa9YqqSneAYQGKNdjXylroc7/JcH7mSpo
> /RvS60l2k4NQ9mJrp8ta8rrwzahqiBLUKBQEs9KA0OqR67uV+p6SSXaVWEqx+819
> bjFqZR1rm2GBoS6xFITZpq1G9eB3RNe7ak7Nzdv5LfgQ4RMbtHvQ+DaSL8bAnyQo
> EMDbe4J7IQGLIQtfmeUnZsz2Lmz02XdYtzIFBc9CKP3EGYJSKDVqAQH4dGwPBkBY
> 9c56ggRXgLBtKcDQijZNM/h6geP6C2fAP/HdLjx0qfydR5ktML66dW6IFN3aAgwW
> lFlQurU8D0Ug0uW/T/0k
> =ogdp
> -END PGP SIGNATURE-
 
 Regards,
 Rohit Yadav
 Software Architect, ShapeBlue
 M. +91 88 262 30892 | rohit.ya...@shapeblue.com
 Blog: bhaisaab.org | Twitter: @_bhaisaab
 
 
 
 Find out more about ShapeBlue and our range of CloudStack related services
 
 IaaS Cloud Design & 
 Build
 CSForge – rapid IaaS deployment framework
 CloudStack Co

Re: Announcing Apache™ CloudStack™ v4.4.3

[Daan Hoogland]

On Tue, Apr 21, 2015 at 10:03 PM, Remi Bergsma  wrote:
> Hi Daan,
>
> Awesome!
>
> Could you have a look at the release notes link? It’s a 404 now.
use http://docs.cloudstack.apache.org/projects/cloudstack-release-notes/en/4.4/

for now, Pierr-Luc pu a ticket out at infra about this. The 4.4.3 tag
in the doc repo didn't get pushed to the github mirror. That's the
root of the issue

Re: SG not working with 6.5 and PR review

[Tim Mackey]

Rohit,

There are no code changes to util.pread2, so what you're observing would
have to be an artifact of either stdout changing, or how popen.communicate
handles the trailing newline from the actual command.  Regardless, your
solution should take care of both cases.

For the ## Discuss, I think we should separate the upgrade from the normal
operation.  For normal operation where we've sustained communication with
the host, then the optimistic case should allow for increased scalability.
In the case of upgrade of management server, host in maintenance mode,
upgrade of host, loss of communication with host, we should assume a very
pessimistic case and upon connection with the host validate the rules,
rebuilding those which don't agree with what we expect.  It's more work,
but validation would help ensure our view of the world is maintained.

-tim

On Tue, Apr 21, 2015 at 3:41 PM, Rohit Yadav 
wrote:

> Hi Tim,
>
> Thanks for your review. I’ve fixed the tick-quote issue on the mentioned
> line. Please find the various issues addressed in this patch;
>
> ## 1. Issue with sm/util.py
>
> My concern with sm/util.py was that, previously on XS 6.2 the output of a
> typical util.pread2().split(‘\n’) would return a list like
> [a,b,c,d,”"] with the last element as an empty string possibly because an
> extra newline in the command output. Due to this all the code in vmops
> plugin followed this kind of patten where the programmer coded it to do a
> list.pop() after calling pread2().split(‘\n’).
>
> But in case of XS 6.5, I found that no newline was at the end and we
> should not assume it and doing a list.pop() would remove the last element
> which can be a valid string and not a newline (or empty string after
> calling split() on it). I traced this issue when I saw following methods
> fail in the /var/log/cloud/cloud.log and looked at the code:
> delete_rules_for_vm_in_bridge_firewall_chain
> destroy_ebtables_rules
> destroy_arptables_rules
>
> The fix I tried was to filter out empty and None elements from the list
> using a "filter(util.pread2().split(‘\n’))” kind of pattern
> instead of doing a list.pop() which might remove a valid string. After this
> fix, I saw that SG rules applied correctly and the previous errors went
> away. I believe that this way of defensive coding would work on both XS 6.2
> and 6.5, irrespective of the output having empty lines or not.
>
>
> ## 2. Issue with large CIDRs and ipset overflow
>
> The other issue I found was when I added an ingress or egress CIDR that
> was in the block /8 or /4. I found that ipset failed in the
> /var/log/cloud/cloud.log. On tracing the issue I found that ipset data
> structure has been modified to use iphash instead of iptreemap (done in
> bd6f03aa954d4b3e7ead7e8010c5674d5d1f9513) because recent versions of ipset
> (as in XS 6.5) did not have that data-structure.
>
> The issue with “iphash” in this case was that when someone added a CIDR or
> CIDRs, it would add individual IPs instead of the CIDR itself. Which failed
> for /8 or /4 because the default max length of “iphash” entry was 65535.
> The fix for this was to use “nethash” since we receive CIDRs only and it
> becomes memory/space efficient. Only the ipset entry for the VM itself is
> “iphash” based now (to store a VM’s primary and secondary IPs).
>
>
> ## 3. Issue with removing Ingress/Egress rules
>
> I saw that when an ingress or egress rule was removed while the iptable
> rules were removed/fixed, the ipset entry stayed there and cidrs in it were
> not getting removed. Every time I added or removed an ingress or egress
> entry, CloudStack would send the vmops plugin all the ingress and egress
> rules. This assumed that all add/remove operations were sort of idempotent
> and all information was applied every time we made any change. Therefore,
> as a pessimistic approach I added the fix to flush/remove the ipset entries
> before adding them back again (that the mgmt server sent to vmops plugin).
> This also solves an upgrade error for users who would upgrade to XS 6.5 or
> ACS 4.5, as previous ipset entries were of type “iptreemap” or “iphash”,
> but new entries (the ingress and egress ones) is of “nethash” type.
>
>
> ## Discuss - should we flush/remove ipset rules before applying new rules?
>
> Can you advise if we should have this pessimistic approach (as it flushes
> ipset entries when rules may be still in place?) or use an optimistic
> approach to keep the entries. There are two issues with the current
> optimistic approach:
> 1. When egress/ingress rules are removed, for a brief period (until new
> ipset entry replace the old one) old entries still persist. But since
> iptables  rules are removed and no one is referencing them it is harmless.
> 2. In case of users upgrading, the old ipset data-structure is “iptreemap”
> or “iphash”, but the new one is “nethash”, how do we make sure that old
> ones are removed/deleted so new ones are of “nethash” type.
>
> > On 21-Apr-2015, at 7:51 pm, Tim 

Re: Next ACS release?

[Andrei Mikhailovsky]

Ilya, Mark, thanks for your feedback, 

I also see the need to restructure the release schedule for ACS as the current 
release cycles are not really working. There is no _reliable_ release cycle of 
the product and as we have recently seen with the 4.5 branch, the release did 
not happen for months and it is still not clear when this will take place. In 
my (I must admit somewhat limited) experience if there are no deadlines, 
developers are not keen on releases and the release are likely to be delayed. 
This is what we've seen with the past ACS releases, they are overdue by many 
months. 

The community might get a much better responce if there is a much shorter 
release cycle even if it means pushing out less features with each release. At 
least some features will get completed, tested and implemented in a set time 
frame. I would rather see a release cycle of every 3-4 months with 5 new 
features than a release with 15 new features which may or may not get released 
every 9 - 12 months. 

By any means, please comment if someone disagrees or thinks there is a better 
alternative. 

Andrei 
- Original Message -

> From: "ilya" 
> To: dev@cloudstack.apache.org
> Sent: Tuesday, 21 April, 2015 7:30:34 PM
> Subject: Re: Next ACS release?

> Andrei,

> To best of my knowledge, both 4.4.x and 4.5.x are being worked on
> actively. As a community, we need to get better on QA of each release
> -
> this is something we are planning to cover this year with distributed
> QA
> model, this was not widely discussed yet but something we need to
> tackle.

> 4.5 rc2 got stalled and we need to restart. We had a 4 month release
> cycle, but we can really stick to it hard - as its community driven.
> May
> will have to revise it down to 6 months or so.

> Regards
> ilya

> On 4/21/15 1:26 AM, Andrei Mikhailovsky wrote:
> > Hello guys,
> >
> > Looking at the dev and user lists it is becoming less certain if
> > version 4.5.x is ever coming out. It seems like a few months have
> > passed since the not so fortunate release of 4.5.0 and I can't
> > find a release schedule for the 4.5.1, which seems to have stopped
> > at rc2 stage and haven't progressed further to a release stage.
> >
> > Are we likely to see any progress with the 4.5.x branch or is the
> > community switching towards the 4.6.x branch without releasing the
> > 4.5.x?
> >
> > I am a bit unclear as there are no release dates, schedules or dead
> > lines that the community should work with. Possibly as a result of
> > this, the ACS releases are not being released on time or fast
> > enough.
> >
> > Does it make sense to introduce release schedules for ACS that the
> > dev community should stick to? Similar to what is being done in
> > many other projects, like Ubuntu, etc. Or would this break the ACS
> > project releases even more?
> >
> > Andrei
> >

[GitHub] cloudstack pull request: CLOUDSTACK-8395: vmops plugin should work...

[agneya2001]

Github user agneya2001 commented on the pull request:

https://github.com/apache/cloudstack/pull/186#issuecomment-95016103
  
Reviewed it, looks good.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---