mirror of
https://git.postgresql.org/git/postgresql.git
synced 2024-09-08 18:39:20 +02:00
dafaa3efb7
Until now, our Serializable mode has in fact been what's called Snapshot Isolation, which allows some anomalies that could not occur in any serialized ordering of the transactions. This patch fixes that using a method called Serializable Snapshot Isolation, based on research papers by Michael J. Cahill (see README-SSI for full references). In Serializable Snapshot Isolation, transactions run like they do in Snapshot Isolation, but a predicate lock manager observes the reads and writes performed and aborts transactions if it detects that an anomaly might occur. This method produces some false positives, ie. it sometimes aborts transactions even though there is no anomaly. To track reads we implement predicate locking, see storage/lmgr/predicate.c. Whenever a tuple is read, a predicate lock is acquired on the tuple. Shared memory is finite, so when a transaction takes many tuple-level locks on a page, the locks are promoted to a single page-level lock, and further to a single relation level lock if necessary. To lock key values with no matching tuple, a sequential scan always takes a relation-level lock, and an index scan acquires a page-level lock that covers the search key, whether or not there are any matching keys at the moment. A predicate lock doesn't conflict with any regular locks or with another predicate locks in the normal sense. They're only used by the predicate lock manager to detect the danger of anomalies. Only serializable transactions participate in predicate locking, so there should be no extra overhead for for other transactions. Predicate locks can't be released at commit, but must be remembered until all the transactions that overlapped with it have completed. That means that we need to remember an unbounded amount of predicate locks, so we apply a lossy but conservative method of tracking locks for committed transactions. If we run short of shared memory, we overflow to a new "pg_serial" SLRU pool. We don't currently allow Serializable transactions in Hot Standby mode. That would be hard, because even read-only transactions can cause anomalies that wouldn't otherwise occur. Serializable isolation mode now means the new fully serializable level. Repeatable Read gives you the old Snapshot Isolation level that we have always had. Kevin Grittner and Dan Ports, reviewed by Jeff Davis, Heikki Linnakangas and Anssi Kääriäinen |
||
---|---|---|
.. | ||
po | ||
.gitignore | ||
common.c | ||
compress_io.c | ||
compress_io.h | ||
dumputils.c | ||
dumputils.h | ||
keywords.c | ||
Makefile | ||
nls.mk | ||
pg_backup_archiver.c | ||
pg_backup_archiver.h | ||
pg_backup_custom.c | ||
pg_backup_db.c | ||
pg_backup_db.h | ||
pg_backup_directory.c | ||
pg_backup_files.c | ||
pg_backup_null.c | ||
pg_backup_tar.c | ||
pg_backup_tar.h | ||
pg_backup.h | ||
pg_dump_sort.c | ||
pg_dump.c | ||
pg_dump.h | ||
pg_dumpall.c | ||
pg_restore.c | ||
README |
src/bin/pg_dump/README Notes on pg_dump ================ 1. pg_dump, by default, still outputs text files. 2. pg_dumpall forces all pg_dump output to be text, since it also outputs text into the same output stream. 3. The plain text output format cannot be used as input into pg_restore. To dump a database into the new custom format, type: pg_dump <db-name> -Fc > <backup-file> or, to dump in TAR format pg_dump <db-name> -Ft > <backup-file> To restore, try To list contents: pg_restore -l <backup-file> | less or to list tables: pg_restore <backup-file> --table | less or to list in a different order pg_restore <backup-file> -l --oid --rearrange | less Once you are happy with the list, just remove the '-l', and an SQL script will be output. You can also dump a listing: pg_restore -l <backup-file> > toc.lis or pg_restore -l <backup-file> -f toc.lis edit it, and rearrange the lines (or delete some): vi toc.lis then use it to restore selected items: pg_restore <backup-file> --use=toc.lis -l | less When you like the list, type pg_restore backup.bck --use=toc.lis > script.sql or, simply: createdb newdbname pg_restore backup.bck --use=toc.lis | psql newdbname TAR === The TAR archive that pg_dump creates currently has a blank username & group for the files, but should be otherwise valid. It also includes a 'restore.sql' script which is there for the benefit of humans. The script is never used by pg_restore. Note: the TAR format archive can only be used as input into pg_restore if it is in TAR form. (ie. you should not extract the files then expect pg_restore to work). You can extract, edit, and tar the files again, and it should work, but the 'toc' file should go at the start, the data files be in the order they are used, and the BLOB files at the end. Philip Warner, 16-Jul-2000 pjw@rhyme.com.au