Re: Infrastructure monitoring
От | John Hansen |
---|---|
Тема | Re: Infrastructure monitoring |
Дата | |
Msg-id | 34F8F335525CB14C95CF92BE2194858E01285E@pdc.geeknet.com.au обсуждение исходный текст |
Ответ на | Infrastructure monitoring ("Jim C. Nasby" <jnasby@pervasive.com>) |
Ответы |
Re: Infrastructure monitoring
|
Список | pgsql-www |
> -----Original Message----- > From: Dave Page [mailto:dpage@vale-housing.co.uk] > Sent: Sunday, January 15, 2006 8:41 AM > To: John Hansen; Joshua D. Drake > Cc: Guido Barosio; Marc G. Fournier; Josh Berkus; > pgsql-www@postgresql.org; Jim C. Nasby > Subject: Re: [pgsql-www] Infrastructure monitoring > > > > > On 14/1/06 21:37, "John Hansen" <john@geeknet.com.au> wrote: > > > Joshua D. Drake [mailto:jd@commandprompt.com] Wrote > > > >> Sure we could help and we are happy to. John what types of > machines > >> are we talking about here? > > > > You will need an x86-32 (Preferably SMP) with at least 2G of ram > > running linux. > > (Gentoo is known to work). > > ASPseek currently uses 11G of HD space for the search. > > What sort of storage are you currently running? How many spindles etc? > > Regards, Dave. > 4x80G Seagate IDE (PATA ST380011A) Software Raid0 on 2 x Promise Ultra100 Controllers in 66Mhz slots. The system has another 4x80 Seagate IDE (PATA ST380021A) but they are slower models, in Software Raid5, used for backup. Pgbench reports ~800tps with: pgbench -i -s50 test; reboot; pgbench -s50 -c25 -t1000 test; (Reboot done to completely flush any cached data) Systained thruput on the disk sybsystem is roughly 130Mbytes/second in both directions. LVM on top of software raid, XFS filesystems. ... John
В списке pgsql-www по дате отправления: