home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!stanford.edu!agate!ucbvax!HPLWK.HPL.HP.COM!albert
- From: albert@HPLWK.HPL.HP.COM (Joseph Albert)
- Newsgroups: comp.databases
- Subject: Re: distributed transactions
- Message-ID: <9208211640.AA07285@hplwk.hpl.hp.com>
- Date: 21 Aug 92 16:40:16 GMT
- Sender: daemon@ucbvax.BERKELEY.EDU
- Lines: 32
-
-
- edwards@world.std.com (Jonathan Edwards) writes:
-
- >>
- >>If a non-blocking commit protocol is used, any transaction started at
- >>a remote site which went down can be aborted, releasing its locks, and
- >>leaving the database in a transaction-consistent state.
- >>
- >Could you site some references or examples for this statement?
- >Certainly the conventional 2-phase commit protocol (which is what is
- >implemented by all commercial instances of distributed TP that I know of)
- >can easily leave you in a blocked-or-inconsistent state.
- >See Bernstein and Goodman: Concurrency Control and Recovery in Database
- >Systems. They state as a 'proposition' that any distributed transaction
- >protocol has to be blocking in the presence of communication failures or
- >total failures.
-
- I don't have a copy of the Bernstein and Goodman book handy, though I thought
- they discussed some sort of non-blocking protocol, the most well-known of
- which is a 3-phase commit. Guess my memory is faulty.
-
- In any case, a good treatment of distributed commit protocols is:
-
- Skeen, D., `Nonblocking Commit Protocols', in Proc. of the ACM-SIGMOD Int'l
- Conference on Management of Data, Orlando, FL, 1981.
-
- Dale Skeen developed a theory of when a protocol is nonblocking and when it
- can block, and further showed how one can generate any one of an infinite
- family of non-blocking commit protocols.
-
- Joseph Albert
- albert@hplabs.hp.com
-