[HBASE-19123] Purge 'complete' support from Coprocesor Observers

• Type: Task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.0.0-beta-1, 2.0.0
• Component/s: Coprocessors
• Hadoop Flags: Reviewed
• Release Note: This issue removes the 'complete' facility that was in ObserverContext. It is no longer possible for a Coprocessor to cut the chain-of-invocation and insist its response prevails.
• Assignee: Michael Stack
• Reporter: Michael Stack
• Created: 30/Oct/17 20:55
• Updated: 21/Mar/18 22:20
• Resolved: 21/Nov/17 06:20

Description

Up on dev list under '[DISCUSSION] Removing the bypass semantic from the Coprocessor APIs', we are discussing purge of 'complete'. Unless objection, lets purge for beta-1.

[e-mail] says the following up on the dev list:

It would simplify the theory of operation for coprocessors if we can assume either the entire chain will complete or one of the coprocessors in the chain will throw an exception that not only terminates processing of the rest of the chain but also the operation in progress.
Security coprocessors interrupt processing by throwing an exception, which is meant to propagate all the way back to the user.
I think it's more than fair to ask the same question about 'complete' as we did about 'bypass': Does anyone use it? Is it needed?

Issue Links

• is related to: HBASE-18770 Remove bypass method in ObserverContext and implement the 'bypass' logic case by case
• relates to: HBASE-19384 Results returned by preAppend hook in a coprocessor are replaced with null from other coprocessor even on bypass
• links to: Review Board (master)