abstractdog commented on PR #305: URL: https://github.com/apache/tez/pull/305#issuecomment-1667337055
I can see that we're doing lot's of interning stuff in the TaskAttemptImpl, which kind of looks like boilerplate from the TaskAttempImpl class' point of view I guess these intern calls were done for a reason: what about wrapping yarn container into a tez implementation that can hide all the interns + null checks? so instead of: ``` StringInterner.intern(RackResolver.resolve(container.getNodeId().getHost()).getNetworkLocation()); ``` let it be: ``` container.getNodeRackName(); ``` basically, a new Container class can have all these fields that you're about to remove from TaskAttemptImpl optionally: look for usages of org.apache.hadoop.yarn.api.records.Container in tez code and remove it from there too (it can be done in followup tickets) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@tez.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
