ddanielr commented on PR #3187: URL: https://github.com/apache/accumulo/pull/3187#issuecomment-1420075671
> > In the client code, the first time the client performs a scan on a server, the client waits for all writes to complete before the scan can begin. It can be argued, that for user tables, the application may not be concerned with waiting for the writes to complete given that the waits are there for a specific use case. > > My understanding is that the specific use case is consistency from the same client, to ensure that a previous write is visible on subsequent scans. The comment says this is especially important for metadata, but I think it's just as important for consistency in general. While this may not be strictly required in all use cases (some users may not care about this consistency), I don't know of a compelling reason to relax the behavior. I don't think effort should be put into making client views of the data less consistent... not without a substantially compelling reason, and I haven't seen any reason even proposed, least of all a substantially compelling one. > > In any case, I definitely don't think we should do anything in 1.10 at all for this. I had asked for branch clarification in #3182. The issue wasn't assigned to a particular branch so I started at 1.10 with the expectation that I could re-target as desired. Happy to move this work to 2.1 instead. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
