Problems using retry() when failures occur in pipeline

[Mike Kingsbury]

I have a fairly simple block of code we run to update a system, that occasionally fails due to problems outside our control.  I thought this would be a great place to wrap it in a retry block.  The problem I have is that even though the retry block is successful later on, it fails the build at the end due to the earlier failure. I also tried unsuccessfully to wrap this in a try/catch without much luck.  Any suggestions on how to not fail a build if the retry block is successful? (I may just pull this out into a separate shell script at this point...)

    int retry_count = 0
    retry(25){
      if ( retry_count != 0 ) {
        sleep(10)
      }
      retry_count++
      sh 'sudo apt -y autoremove'

      sh 'sudo apt-get update'
      sh 'sudo apt-get -y -f upgrade'
      sh './make-deps.sh'
    }