Sounds good. I can file a ticket for Hadoop 3 support. But for "you should lay out your plan how to attack the issue", I don't really know what is needed except making it build and work. Could you give me some pointers?
What is the process of creating a new branch? I've never done that before. On Sat, Oct 20, 2018 at 3:27 AM Olaf Flebbe <[email protected]> wrote: > Hi Dagang, > > There were some voices asking for Hadoop-3 support but nobody stepped up > to commit something. > > Please create JIRA tickets if you like to work on it, community surely > will appreciate. Anybody can create tickets, make pull requests or submit > patches. Bigtop committers will review your contributions. > > I would recommend to have hadoop-3 handled on a separate branch, since it > will break current Bigtop. > > Maybe you should lay out your plan how to attack the issue, in order to > make it easier to follow your work. > > I did some initial investigations (still visible in github branch > "bigtop-alpha") but I you may like to have a fresh start to handle hadoop-3. > > Olaf > > > > > > Am 19.10.2018 um 18:20 schrieb Dagang Wei <[email protected]>: > > > > Yes, I am looking to support Hadoop 3 (internally)? Is there plan to > > support Hadoop 3 in Bigtop? If we have a tracking ticket, I might be able > > to work on it. > > > > On Thu, Oct 18, 2018 at 7:30 PM Evans Ye <[email protected]> wrote: > > > >> I'm also not an expert in Bigtop packaging. But to support Hadoop 3 I > think > >> we have a great chance to overhaul the hadoop packaging in Bigtop. Are > you > >> planning to add support of Hadoop 3? > >> > >> Dagang Wei <[email protected]> 於 2018年10月11日 週四 上午6:40寫道： > >> > >>> Hi folks, > >>> > >>> I have a question regarding installing scripts under > >>> /usr/lib/hadoop/libexec/ in [hadoop.install]( > >>> > >>> > >> > https://github.com/apache/bigtop/blob/master/bigtop-packages/src/deb/hadoop/hadoop.install > >>> ): > >>> > >>> /usr/lib/hadoop/libexec/hadoop-config.sh > >>> /usr/lib/hadoop/libexec/hadoop-layout.sh > >>> > >>> Why do we specify hadoop-config.sh and hadoop-layout.sh individually? > why > >>> not simply install the whole directory by simply > >>> specify /usr/lib/hadoop/libexec? Because I think that would be more > >>> flexible, in Hadoop3 new scripts e.g., hadoop-functions.sh are > >> introduced. > >>> > >>> Thanks! > >>> Dagang > >>> > >> > >
