Thread: 2Q implementaion for PostgreSQL buffer replacement.
Hi. I implement 2Q algorithm to PostgreSQL for buffer management , instead of LRU. It's known as low overhead and high performance than LRU. If you have some interests , see following URL. http://www.vldb.org/conf/1994/P439.PDF In my test (pgbench -S) , it improves 4% cache hit rate and 2% up performance comparing from LRU. Do you have any interest about this patch? -- Yutaka tanida <yutaka@nonsensecorner.com> http://www.nonsensecorner.com/
Attachment
Looks good to me --- we will include it in 7.4. Your patch has been added to the PostgreSQL unapplied patches list at: http://momjian.postgresql.org/cgi-bin/pgpatches I will try to apply it within the next 48 hours. --------------------------------------------------------------------------- Yutaka tanida wrote: > Hi. > > I implement 2Q algorithm to PostgreSQL for buffer management , instead > of LRU. > It's known as low overhead and high performance than LRU. If you have > some interests , see following URL. > > http://www.vldb.org/conf/1994/P439.PDF > > In my test (pgbench -S) , it improves 4% cache hit rate and 2% up > performance comparing from LRU. > > Do you have any interest about this patch? > > -- > Yutaka tanida <yutaka@nonsensecorner.com> > http://www.nonsensecorner.com/ [ Attachment, skipping... ] [ Attachment, skipping... ] [ Attachment, skipping... ] > > ---------------------------(end of broadcast)--------------------------- > TIP 4: Don't 'kill -9' the postmaster -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
On Mon, 23 Jun 2003 23:49:17 -0400 (EDT) Bruce Momjian <pgman@candle.pha.pa.us> wrote: > > Looks good to me --- we will include it in 7.4. Thanks.But please note it is not completed yet. I must implement more , and move configurable parameter to postgresql.conf file. -- Yutaka tanida <yutaka@nonsensecorner.com> http://www.nonsensecorner.com/
OK, thanks. I will remove it from the queue, and someone suggested a different algorithm today: > I was researching on cache replacement strategy as well. 2Q has one > disadvantage see this exellent paper: > http://www.almaden.ibm.com/cs/people/dmodha/#ARC see the paper > "ARC: A Self-Tuning, Low Overhead Replacement Cache" for theory and "One > Up on LRU" for implementation details. ARC requires no tuning and can > switch fast between chaging patterns. Best of all is it is resistant to a > "sequential scan" pattern. and i think it's even easier to implement then > 2q :) --------------------------------------------------------------------------- Yutaka tanida wrote: > > On Mon, 23 Jun 2003 23:49:17 -0400 (EDT) > Bruce Momjian <pgman@candle.pha.pa.us> wrote: > > > > > Looks good to me --- we will include it in 7.4. > > Thanks.But please note it is not completed yet. I must implement more , > and move configurable parameter to postgresql.conf file. > > -- > Yutaka tanida <yutaka@nonsensecorner.com> > http://www.nonsensecorner.com/ > > > ---------------------------(end of broadcast)--------------------------- > TIP 9: the planner will ignore your desire to choose an index scan if your > joining column's datatypes do not match > -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
Patch removed at author's request. --------------------------------------------------------------------------- Yutaka tanida wrote: > Hi. > > I implement 2Q algorithm to PostgreSQL for buffer management , instead > of LRU. > It's known as low overhead and high performance than LRU. If you have > some interests , see following URL. > > http://www.vldb.org/conf/1994/P439.PDF > > In my test (pgbench -S) , it improves 4% cache hit rate and 2% up > performance comparing from LRU. > > Do you have any interest about this patch? > > -- > Yutaka tanida <yutaka@nonsensecorner.com> > http://www.nonsensecorner.com/ [ Attachment, skipping... ] [ Attachment, skipping... ] [ Attachment, skipping... ] > > ---------------------------(end of broadcast)--------------------------- > TIP 4: Don't 'kill -9' the postmaster -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073