Bruce Momjian <[EMAIL PROTECTED]> writes: > Is there significant time between entry of CommitTransaction() and the > fsync()? Maybe not. I doubt it. No I/O anymore, anyway, unless the commit record happens to overrun an xlog block boundary. regards, tom lane
- [HACKERS] CommitDelay performance improvement Tom Lane
- Re: [HACKERS] CommitDelay performance improvement Bruce Momjian
- Re: [HACKERS] CommitDelay performance improvement Tom Lane
- Re: [HACKERS] CommitDelay performance improveme... Bruce Momjian
- Re: [HACKERS] CommitDelay performance impro... Tom Lane
- Re: [HACKERS] CommitDelay performance ... Bruce Momjian
- Re: [HACKERS] CommitDelay performance improvement Nathan Myers
- Re: [HACKERS] CommitDelay performance improvement Bruce Momjian
- Re: [HACKERS] CommitDelay performance improvement Tom Lane
- Re: [HACKERS] CommitDelay performance improveme... Bruce Momjian
- Re: [HACKERS] CommitDelay performance impro... Tom Lane
- Re: [HACKERS] CommitDelay performance ... Bruce Momjian
- Re: [HACKERS] CommitDelay performa... Tom Lane
- Re: [HACKERS] CommitDelay perf... Bruce Momjian
- Re: [HACKERS] CommitDelay perf... Tom Lane