Anonymous Modules "runtime context"

Happy New Year, everyone! During the break, I've been playing with a toy anon module[0] mostly for "learning" purposes. In doing so, I realized it would be useful, as a developer, to get access to even a "minimal" runtime context and filed MESOS-4253[1]. (there was also a TODO from benh in the

Re: GPGPU SLA support

I have posted this on the user list. Reposting here for continuity of this thread. I have been working on GPU QoS by modifying the GPU kernels. Something along the lines of this . The basic mechanism is to time slice a grid by running a

Re: mesos git commit: Added documentation for API versioning.

Maybe I'm nit-picking, but I thought you mean only API versioning, which does not include release versioning, to which "Updates" section is related. I would propose to make the sentence more catchy: convert it to `NOTE` and state clear the whole doc (not just API versioning) applies to Mesos 1.0+

Looking for a shepherd

Hello everyone, My name is Jocelyn and I'm working on a mesos framework for a big classified website in France. Last week when I tried to build mesos 0.26 I encountered an error when compiling libprocess configured with the --disable-bundled. The bug report is here:

Re: mesos git commit: Added documentation for API versioning.

The "Upgrades" section says "The master and agent are typically compatible as long as they are running the same major version.". Is my understanding correct that this will apply to 1.0.0+ versions? Currently we advice people not to skip any Mesos releases, could you please clarify that in the doc?

Re: Fix for MESOS-3809 not in 0.26.0 (Was: Re: [RESULT][VOTE] Release Apache Mesos 0.26.0) (rc5)

This is already taken care of. See https://issues.apache.org/jira/browse/MESOS-3809. This jira will be cherry-picked into 0.27. On Mon, Jan 4, 2016 at 2:47 PM, Benjamin Mahler wrote: > +0.26.0 release managers > > On Sun, Jan 3, 2016 at 12:34 PM, Anindya Sinha

Change minimum supported version for GCC to 4.8.1

I would like to propose that we bump our minimum supported version for gcc from 4.8.0 to 4.8.1. The main motivation behind this is that there are at least 2 outstanding reviews on RB that want to use ref-qualifiers introduced

Re: Change minimum supported version for GCC to 4.8.1

+1 (binding) I would like to propose that we bump our minimum supported version for gcc > from 4.8.0 to 4.8.1. The main motivation behind this is that there are at > least 2 outstanding reviews on RB that want to use ref-qualifiers < >

Re: Change minimum supported version for GCC to 4.8.1

+1 On Mon, Jan 4, 2016 at 3:19 PM, Joris Van Remoortere < joris.van.remoort...@gmail.com> wrote: > +1 (binding) > > I would like to propose that we bump our minimum supported version for gcc > > from 4.8.0 to 4.8.1. The main motivation behind this is that there are at > > least 2 outstanding

Re: tasks not being scheduled; cfs_rq for /mesos is missing

hi Jojy, Unfortunately, I haven't been able to reproduce this issue on demand, it has just happened spontaneously a few times. So I cannot say for sure if it would happen on a newer mesos/kernel version. I'm thinking of trying to force reproduction by creating and destroying a ton of cgroups,

Re: Change minimum supported version for GCC to 4.8.1

When we moved to 4.8 there was a spreadsheet that showed how folks can get 4.8 on various distributions, have you checked that 4.8.1 is available across distributions? On Mon, Jan 4, 2016 at 4:07 PM, Adam Bordelon wrote: > +1 > > On Mon, Jan 4, 2016 at 3:19 PM, Joris Van

Scheduler::executorLost() callback will be implemented in 0.27

Dear framework owners, Scheduler::executorLost() callback will be implemented in the future 0.27 release. In previous versions, this callback in the scheduler interface was never called. You could use it to perform

looking for Vamp canary-testing feedback

Hello everybody, first i want to wish you and your families all the best for 2016! I have a small request, i have no intention to offend anyone on this list, as i think it’s topic-specific and suited to a lot of the mesos use-cases we’re hearing. We are building Vamp (www.vamp.io

Re: Change minimum supported version for GCC to 4.8.1

Are you referring to the spread sheet linked in MESOS-2604 ? AFAICT, it just shows that a particular variant of GCC 4.8+ is available on each of the supported distributions. So, this should not be an issue unless I am missing something? -anand

Re: tasks not being scheduled; cfs_rq for /mesos is missing

Hi Eric, Thanks again for the information. The value of 0 for */cgroup/cpu/cpu.cfs_* files looks suspicious. Since you are motivated to look into kernel, here is what I think is interesting. - The shed_debug print

Re: Change minimum supported version for GCC to 4.8.1

+1 On Mon, Jan 4, 2016 at 9:36 PM Anand Mazumdar wrote: > Are you referring to the spread sheet linked in MESOS-2604 < > https://issues.apache.org/jira/browse/MESOS-2604>? > > AFAICT, it just shows that a particular variant of GCC 4.8+ is available > on each of the

Re: Fix for MESOS-3809 not in 0.26.0 (Was: Re: [RESULT][VOTE] Release Apache Mesos 0.26.0) (rc5)

+0.26.0 release managers On Sun, Jan 3, 2016 at 12:34 PM, Anindya Sinha wrote: > Looks like the fix for MESOS-3809 did not make into this release 0.26.0 > (based on the tag 0.26.0, or 0.26.0-rc2 when it was first supposed to have > been added) although the CHANGELOG

Re: Looking for a shepherd

Glad to hear you're using mesos! I introduced this bug, sorry! I just pushed your fix, thanks for sending it! On Mon, Jan 4, 2016 at 3:08 AM, Jocelyn De La Rosa < jocelyn.delar...@scmfrance.fr> wrote: > Hello everyone, > > My name is Jocelyn and I'm working on a mesos framework for a big >

Re: mesos git commit: Added documentation for API versioning.

Vinod alluded to a similar concern as a review comment and we added the following to the beginning of the doc: “API versioning was introduced in Mesos 0.24.0 and this scheme only applies to Mesos 1.0.0 and higher.” I was under the impression that the above statement should suffice? -anand >