0

I'm trying to install hadoop-tools on my system from GitHub: https://github.com/rjoberon/hadoop-tools but my problem is that when I'm trying to run mvn install it is trying to download some dependencies, but I don't have an internet connection for that Hadoop system. On my system, I have all the required dependencies packages in a location. How can I indicate maven to install the dependencies and also the package for Hadoop-tools from that certain location? On the main pom file, I have the repository for Cloudera, and I don't know if I can change this..

bicanul123
  • 427
  • 7
  • 21
  • You could mirror maven central into a local system that does have internet, then configure the hadoop nodes to use that – OneCricketeer Jan 18 '21 at 15:19
  • @OneCricketeer I can store jar files and .pom files for each jar on a server, but I don't know how can I point to do this, cause I have over 500 jars... On the main .pom file I have the address for cloudera repository – bicanul123 Jan 19 '21 at 07:19
  • Hadoop tools generally have many dependencies, yes. You can individually install them in the duplicate, or you can change your Maven settings to use a different repository that is in your local network and acts like a proxy (Artifactory or Sonatype Nexus) – OneCricketeer Jan 19 '21 at 15:13
  • @OneCricketeer so basically there is no other easy way to do this.. I was thinking if I'm copying the .m2 folder from another machine, it would work... – bicanul123 Jan 20 '21 at 10:18
  • That should work, too, considering JARs are cross-platform, though I've ran into issues in the past where JARs get manipulated/corrupted by anti-virus scanners, so you don't have some verifcation process like a Maven Mirror would provide – OneCricketeer Jan 20 '21 at 18:08
  • @OneCricketeer I've copied it, but I got some weird errors that I somehow have some missing modules :( Sometimes I get an index out of bound exception for some java files, but the error it's not too explicit... – bicanul123 Jan 21 '21 at 07:41

0 Answers0