Happy to report as of the last two PRS, FGS is usable no memory leaks or
crashes, could likely be improved with fancier schedulers but that's for
the future. I'm currently looking at running some terasort benchmarks with
FGS and a reserved resources vs statically sized NMs to figure out the
Adam:
I agree with your suggestion for 0.2. I am, however, curious to understand
why a dockerized NM would be a requirement for multi-tenancy. I don't
disagree, I'm just trying to understand.
paul
On Wed, Mar 16, 2016 at 8:48 PM, Adam Bordelon wrote:
> +1 on Darin as
+1 Darin for RM.
Would like be involved in the multi-tenency with Dockerized NM work.
Perhaps we can start filing JIRA's targeted for 0.2 and start adding design
proposals
for people to review and comment.
Regards
Swapnil
On Thu, Mar 17, 2016 at 8:13 AM, Darin Johnson
Hi Darin,
container-executor.cfg on the master node:
==
yarn.nodemanager.linux-container-executor.group=yarn #configured value
of yarn.nodemanager.linux-container-executor.group
banned.users=#comma separated list of users who can not run
Hi Darin,
thanks a lot for this. But what about the other case below, when cgroups
is disabled?
Björn
Am 18.03.2016 um 00:25 schrieb Darin Johnson:
> Hey Bjorn,
>
> I think I figured out the issue. Some of the values for cgroups are still
> hardcoded in myriad. I'll add a JIRA Ticket
+1 on Darin as release manager
I'd like to see 0.2 have:
- Usable FGS
- Dockerized NM (for multitenancy)
On Tue, Mar 15, 2016 at 9:46 AM, Darin Johnson
wrote:
> We've talked about a 0.2.0 release slated for mid April at the dev sync.
> I'd like to nail down any