I also agree this is minor and did not intend to re-roll. My question is whether CASSANDRA-12758 should go to to the 'cassandra-2.1' branch and be tagged with fixver of '2.1.x' in JIRA? Is this minor improvement satisfactory for a the critical-only nature of the 2.1 branch and go into the next 2.1 release, or leave it for 2.2+?
-- Kind regards, Michael On 10/10/2016 03:09 PM, Nate McCall wrote: >> It's too minor for a re-roll, and safe enough to just apply yourself if you >> want it. > > Agreed. > >> >> On Mon, Oct 10, 2016 at 2:44 PM, Michael Shuler <mich...@pbandjelly.org> >> wrote: >> >>> Nate, do think CASSANDRA-12758 should go to 2.1.x? >>> >>> -- >>> Michael >>> >>> On 10/10/2016 02:26 PM, Nate McCall wrote: >>>> Hi Romain, >>>> I appreciate you speaking up about this, but I stuck with my +1 in >>>> order to get 2.1.16 with the NTR fix out since I have seen >>>> CASSANDRA-11363 with every recent client installation. Also, running >>>> the patch in production produced results satisfactory enough to me to >>>> preclude the need for explicit monitoring added by your patch (though >>>> I do think it's a good idea to have a metric). >>>> >>>> Thanks for both the patch and bringing it up regardless. >>>> >>>> -Nate >>>> >>>> On Fri, Oct 7, 2016 at 11:45 AM, Romain Hardouin >>>> <romainh...@yahoo.fr.invalid> wrote: >>>>> Hi, >>>>> I use the "current 2.1.16" (commit >>>>> cdd535fcac4ba79bb371e8373c6504d9e3978853) >>> on production in 5 DCs (82 nodes) out of 7 and it works well!I've just had >>> to add a MBean to track changes of the NTR queue length on top of cdd535f. >>> This allow to make correlations with other metrics and see the impact of a >>> change. >>>>> I've filed a ticket with patches for 2.1 and the trunk >>> https://issues.apache.org/jira/browse/CASSANDRA-12758 >>>>> Do you think this MBean could land in the final 2.1.16 since it goes >>> hand-in-hand with CASSANDRA-11363? >>>>> >>>>> Thanks, >>>>> Romain >>> >>>