home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.databases.sybase
- Path: sparky!uunet!shearson.com!newshost!wfinnert
- From: wfinnert@larry.shearson.com (Warren Finnerty)
- Subject: Re: recovery time at start up
- In-Reply-To: chris@cs.uchicago.edu's message of 30 Dec 92 19:31:44 GMT
- Message-ID: <WFINNERT.92Dec31111928@larry.shearson.com>
- Sender: news@shearson.com (News)
- Organization: Lehman Brothers
- References: <1992Dec28.181641.4653@trdlnk.uucp>
- <1992Dec30.193144.26562@midway.uchicago.edu>
- Date: Thu, 31 Dec 1992 16:19:28 GMT
- Lines: 22
-
- In article <1992Dec30.193144.26562@midway.uchicago.edu> chris@cs.uchicago.edu (Chris Johnston) writes:
-
- > My server will recover quickly if I checkpoint my databases
- > before I shutdown. A checkpoint only takes one second and avoids
- > several minutes of recovery time.
-
- > I try to checkpoint my databases often, others on this list have
- > warned against doing it too often. I don't know why.
-
-
- - A "shutdown" will checkpoint all the databases before shutting down.
- - "shutdown with nowait" and UNIX "kill" will not.
- - No harm in checkpointing often.
- - A long recovery is usually related to rolling *back* a large transaction
- the max size of which is governed by the transaction log.
-
-
- --
- warren finnerty | 388 Greenwich St.
- Lehman Brothers | NYC NY 10013
- "Back off man!" | wfinnert@shearson.com
-
-