Conversation

Notices

1. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 12:40:09 EDT musicman
   back on the #Ambari train this morning, with CentOS 7
   1. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 12:41:42 EDT musicman

      in reply to
      wtf is going on with the duplicates...
      1. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 12:45:05 EDT musicman

         in reply to
         "resolved"
   2. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 12:57:06 EDT musicman

      in reply to
      unsurprisingly, #maven version so old in CentOS that it is unusable. maven now compiled from source, and doing its own compilation.
      1. lnxw48a1 (lnxw48a1@nu.federati.net)'s status on Thursday, 03-Sep-2020 13:19:29 EDT lnxw48a1

         in reply to
         @musicman Yeah, RHEL / CentOS is about stability, not newness. That said, is there a reason you're still running 7.x instead of 8.x? Upgrading from 6.5 to 7.0 a few years back meant a lot of EOL tweaks (such as add-on repos that had a currently supported version of PHP) were unnecessary.
         1. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 13:34:38 EDT musicman

            in reply to
            yes, Ambari does not support CentOS 8. CentOS 7 doesn't seem to be going much better though.
         2. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 13:35:53 EDT musicman

            in reply to
            misread "Your node version must <= v11.15.0" initially, but doesn't seem to matter what version of node is installed
         3. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 13:40:46 EDT musicman

            in reply to
            I haven't tried modifying the pom.xml directly yet, which is what the PR suggests. The PR hasn't been merged yet and all the tests seemed to have failed, so I can't say I have high hopes.
            
            Still, testing environment might have limitations my environment does not
            1. lnxw48a1 (lnxw48a1@nu.federati.net)'s status on Thursday, 03-Sep-2020 14:26:32 EDT lnxw48a1

               in reply to
               @musicman You can modify pomade.xml, which describes a hairstyle, to see if that helps.
      2. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 13:32:29 EDT musicman

         in reply to
         well, we shall see: https://issues.apache.org/jira/browse/AMBARI-25540
   3. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 14:44:11 EDT musicman

      in reply to
      the ambari install is garbage: https://stackoverflow.com/questions/63085657/build-ambari-2-7-5-on-centos-7-8-from-sources-error
   4. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 14:48:32 EDT musicman

      in reply to
      and another error: https://www.mail-archive.com/user@ambari.apache.org/msg03473.html
   5. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 14:56:29 EDT musicman

      in reply to
      on the plus side, I am getting very comfortable editing maven files, and I even learned something super handy about vim
   6. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 16:08:30 EDT musicman

      in reply to
      um, why would you ship this?
      
      <!-- ToDo: remove after moving the Metrics HDP dependencies to the public nexus -->
      <repository>
      <id>private-nexus</id>
      <name>private-nexus</name>
      <url>https://nexus-private.hortonworks.com/nexus/content/groups/public/</url>
      <snapshots>
      <enabled>true</enabled>
      </snapshots>
      <releases>
      <enabled>true</enabled>
      </releases>
      </repository>
      <!-- //ToDo: END -->
      1. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 16:10:18 EDT musicman

         in reply to
         I wonder if going back to 2.7.4 would fix it, or if I am just going to have to chase down whatever this "publc nexus" is
      2. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 16:34:39 EDT musicman

         in reply to
         I mean, it would be one thing if it worked...
   7. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 16:26:47 EDT musicman

      in reply to
      I figured out I was looking in the wrong place for the hbase stuff. It was in org.apache.hbase. #java is so weird.
   8. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 16:33:42 EDT musicman

      in reply to
      I'm going to try just skipping the metrics. I'm tired of messing with it. This was supposed to make Kafka and Cassandra stuff easier...
   9. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 16:40:53 EDT musicman

      in reply to
      skipping the upper level didn't seem to work. Idk if it worked at all, or just failed to cascade. This time trying to skip the exact module. Maybe the rest of the metrics will work
   10. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 17:02:37 EDT musicman

       in reply to
       might be getting somewhere...
       
       https://mail-archives.apache.org/mod_mbox/ambari-user/202005.mbox/browser
   11. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 17:06:16 EDT musicman

       in reply to
       so...it seems like Apache BigTop might be the way to go
       1. lnxw48a1 (lnxw48a1@nu.federati.net)'s status on Thursday, 03-Sep-2020 17:09:07 EDT lnxw48a1

          in reply to
          @musicman Thanks. I was not aware of https://bigtop.apache.org/ before now.
       2. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 17:13:43 EDT musicman

          in reply to
          there are still commits, but https://github.com/apache/bigtop/releases is not a good look
       3. musicman (musicman@nu.federati.net)'s status on Thursday, 03-Sep-2020 17:20:25 EDT musicman

          in reply to
          we don't support BigTop and the install instructions leave a lot to be desired, so I am just going to give up and do Kafka and Cassandra "the old fashioned way"