Hi Xinyu, How about we move the gRPC zero-copy code to a feature branch first [1]? Please see if you could review the master-tmp branch [2]. If it is good, I will continue merging the commits after RATIS-2269, which the last commit in master-tmp.
Tsz-Wo [1] https://lists.apache.org/thread/pmfjl4kyfkzbtcpdznbjrjv6gr4844hf [2] https://github.com/szetszwo/ratis/tree/master-tmp On Tue, May 13, 2025 at 8:26 PM Xinyu Tan <[email protected]> wrote: > Hi， > > We did not find any problems with the snapshot version [1] on master, and > I see that the thirdparty[2] update has been merged at the moment. If there > are no problems, can we just release 3.2.0 based on the master branch? > > [1] > https://repository.apache.org/content/groups/snapshots/org/apache/ratis/ratis-server/3.2.0-3247c7f-SNAPSHOT/ > [2] https://github.com/apache/ratis/pull/1252 > > Best > ------------- > Xinyu Tan > > On 2025/04/28 18:55:06 Tsz-Wo Nicholas Sze wrote: > > > ... Hope it can be completed before the May Day holiday. > > > > Do you mean the first week of May? If you are in a hurry, please feel > free > > to roll a release without waiting for anything. We could roll another > > release (say 3.2.0) afterward. > > > > > Do you have any more recommended commits for 3.1.4? > > > > For 3.1.4, how about we first remove the gRPC zero-copy from the master > [1] > > ? Just have checked the master-tmp branch [2]. It seems that all the > > commits should be included in 3.1.4. What do you think? > > > > Tsz-Wo > > > > [1] https://lists.apache.org/thread/27b2518ytpbkwknfpg6orzqd9vx1olxt > > [2] https://github.com/szetszwo/ratis/tree/master-tmp > > > > > > On Sun, Apr 27, 2025 at 2:52 AM Xinyu Tan <[email protected]> wrote: > > > > > Hi, Tsz-Wo > > > > > > Thanks a lot! We will wait for the thirdparty to release. Hope it can > be > > > completed before the May Day holiday. > > > > > > Do you have any more recommended commits for 3.1.4? > > > > > > Best > > > ------------- > > > Xinyu Tan > > > > > > On 2025/04/25 04:31:19 Tsz-Wo Nicholas Sze wrote: > > > > Hi Xinyu, > > > > > > > > Thanks a lot for bringing up the next Ratis release! > > > > > > > > For Ratis 3.1.3, the slowness was due to the underlying gRPC and > Netty > > > > versions. We have tested various gRPC and Netty version > combinations. > > > It > > > > seems that we have found some combinations work well. There is a > > > > ratis-thirdparty release VOTE going on [1]. > > > > > > > > For Ratis 3.1.4 (the π version), it is probably good to wait for the > > > > ratis-thirdparty VOTE. Unfortunately, it currently only has two > binding > > > > votes. Let me ping the other PMC members to cast their vote. > > > > > > > > BTW, thanks RkGrit for fixing the stability issues! > > > > > > > > Tsz-Wo > > > > [1] https://lists.apache.org/thread/mp19fwh7l53f5fk4t72ls3093t13rkoq > > > > > > > > > > > > > > > > On Thu, Apr 24, 2025 at 8:35 PM Wei-Chiu Chuang <[email protected]> > > > wrote: > > > > > > > > > Attila found Ozone MapReduce tests were much slower using Ratis > 3.1.3 > > > > > HDDS-12103 <https://issues.apache.org/jira/browse/HDDS-12103> > > > > > Was the regression fixed in 3.1.4? > > > > > > > > > > Thanks > > > > > > > > > > On Thu, Apr 24, 2025 at 8:19 PM Xinyu Tan <[email protected]> > wrote: > > > > > > > > > > > Hi all, > > > > > > > > > > > > Recently, we’ve fixed a series of stability issues related to > Ratis > > > > > > membership changes, including 6 commits from RkGrit. > > > > > > > > > > > > Now, the new release of IoTDB would like to depend on a Ratis > release > > > > > that > > > > > > includes these fixes. I’d like to know if community has > suggestions > > > or > > > > > > feedback if I take on the role of release manager to publish > version > > > > > 3.1.4, > > > > > > mainly consisting of these small bug fixes. > > > > > > > > > > > > If there are no major concerns, I plan to start the voting > process > > > before > > > > > > the May Day holiday in China. > > > > > > > > > > > > Best > > > > > > -------------- > > > > > > Xinyu Tan > > > > > > > > > > > > > > > > > > > > >
