Re: Big directory size
Yes, the kernel has a dirhash option. Thank you for the answer. Alexander Varshavchick, Metrocom Joint Stock Company Phone: (812)118-3322, 118-3115(fax) On 13 Jan 2003, Lowell Gilbert wrote: > Date: 13 Jan 2003 16:36:06 -0500 > From: Lowell Gilbert <[EMAIL PROTECTED]> > To: [EMAIL PROTECTED] > Subject: Re: Big directory size > > Varshavchick Alexander <[EMAIL PROTECTED]> writes: > > > I had a directory with a lot of files (about 100 000), and naturally, the > > size of the directory entry itself was big enough (about 1M). Now I've > > split all these files to different subdirectories, to increase the system > > performance. The major directory entry size didn't change, however such a > > big value is not needed now. > > > > Now here is a question - can an unnessesary big value of a directory > > entry size harm the system performance? I guess it does not, is it > > correct? > > If the files were created without the dirhash code in your kernel, it > certainly could. It still could with the dirhash, but shouldn't be > noticeable at the 100,000-file level. > > To Unsubscribe: send mail to [EMAIL PROTECTED] > with "unsubscribe freebsd-questions" in the body of the message > To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
Re: Big directory size
Yes, talking about a directory entry size, I meant just the size of inode entries, not the summary size of directory contents... Alexander Varshavchick, Metrocom Joint Stock Company Phone: (812)118-3322, 118-3115(fax) On Mon, 13 Jan 2003, Daniel Bye wrote: > Date: Mon, 13 Jan 2003 17:28:50 + > From: Daniel Bye <[EMAIL PROTECTED]> > To: [EMAIL PROTECTED] > Subject: Re: Big directory size > > On Mon, Jan 13, 2003 at 02:45:16PM +0300, Varshavchick Alexander wrote: > > Hello, > > > > I had a directory with a lot of files (about 100 000), and naturally, the > > size of the directory entry itself was big enough (about 1M). Now I've > > split all these files to different subdirectories, to increase the system > > performance. The major directory entry size didn't change, however such a > > big value is not needed now. > > > > Now here is a question - can an unnessesary big value of a directory > > entry size harm the system performance? I guess it does not, is it > > correct? > > This is just speculation, but I would imagine the key factor affecting > performance would be more to do with the number of inode entries in a > given directory, than with the size of the directory's contents. However, > I am no file system expert, this is just my gut feeling... > > -- > Daniel Bye > > PGP Key: ftp://ftp.slightlystrange.org/pgpkey/dan.asc > PGP Key fingerprint: 3D73 AF47 D448 C5CA 88B4 0DCF 849C 1C33 3C48 2CDC > _ > ASCII ribbon campaign ( ) > - against HTML, vCards and X > - proprietary attachments in e-mail / \ > > To Unsubscribe: send mail to [EMAIL PROTECTED] > with "unsubscribe freebsd-questions" in the body of the message > To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
Re: Big directory size
Varshavchick Alexander <[EMAIL PROTECTED]> writes: > I had a directory with a lot of files (about 100 000), and naturally, the > size of the directory entry itself was big enough (about 1M). Now I've > split all these files to different subdirectories, to increase the system > performance. The major directory entry size didn't change, however such a > big value is not needed now. > > Now here is a question - can an unnessesary big value of a directory > entry size harm the system performance? I guess it does not, is it > correct? If the files were created without the dirhash code in your kernel, it certainly could. It still could with the dirhash, but shouldn't be noticeable at the 100,000-file level. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
Re: Big directory size
On Mon, Jan 13, 2003 at 02:45:16PM +0300, Varshavchick Alexander wrote: > Hello, > > I had a directory with a lot of files (about 100 000), and naturally, the > size of the directory entry itself was big enough (about 1M). Now I've > split all these files to different subdirectories, to increase the system > performance. The major directory entry size didn't change, however such a > big value is not needed now. > > Now here is a question - can an unnessesary big value of a directory > entry size harm the system performance? I guess it does not, is it > correct? This is just speculation, but I would imagine the key factor affecting performance would be more to do with the number of inode entries in a given directory, than with the size of the directory's contents. However, I am no file system expert, this is just my gut feeling... -- Daniel Bye PGP Key: ftp://ftp.slightlystrange.org/pgpkey/dan.asc PGP Key fingerprint: 3D73 AF47 D448 C5CA 88B4 0DCF 849C 1C33 3C48 2CDC _ ASCII ribbon campaign ( ) - against HTML, vCards and X - proprietary attachments in e-mail / \ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
Big directory size
Hello, I had a directory with a lot of files (about 100 000), and naturally, the size of the directory entry itself was big enough (about 1M). Now I've split all these files to different subdirectories, to increase the system performance. The major directory entry size didn't change, however such a big value is not needed now. Now here is a question - can an unnessesary big value of a directory entry size harm the system performance? I guess it does not, is it correct? Thanks Alexander Varshavchick, Metrocom Joint Stock Company Phone: (812)118-3322, 118-3115(fax) To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message
