How large file is really large - pathconf results
Regarding to discussion about large segment size of table files a test
pathconf function (see
http://www.opengroup.org/onlinepubs/009695399/functions/pathconf.html).
You can see output there:
_PC_FILESIZEBITS - 3rd column
_PC_LINK_MAX - 4th column
_PC_NAME_MAX - 5th column
_PC_PATH_MAX - 6th column
Solaris Nevada ZFS 64 -1 255 1024
UFS 41 32767 255 1024
FAT 33 1 8 1024
NFS 41 32767 255 1024
Solaris 8 UFS 41 32767 255 1024
NFS 40 32767 255 1024
Centos4(2.6.11) EXT3 64 32000 255 4096
XFS 64 2147483647 255 4096
Mac OSX leopard HFS+ 64 32767 255 1024
The result is not really good :(. I tested it also on HP.UX 11.11/11.23,
Tru64 v4.0 and MacOS tiger (big thanks to Tomas Honzak for machine
access) and Tiger and Tru64 does not recognize _PC_FILESIZEBITS
definition and HP_UX returns errno=EINVAL. I also don't trust Linux
result on EXT3. It seems that only Solaris and Leopard returns
relatively correct result (33 bit on FAT FS is probably not correct).
I attached my test program, please let me know your result from your
favorite OS/FS (binary must be saved on tested FS).
However, I think we cannot use this method to test max file size on FS :(.
Comments, ideas?
Zdenek
PS: Does pg_dump strip a large file or not?
Attachments:
pathconf.ctext/x-csrc; name=pathconf.cDownload
#include <unistd.h>
#include <stdio.h>
#include <errno.h>
#include <string.h>
#include <limits.h>
int main(int argc, char** argv)
{
long ret;
int err;
errno = 0;
ret = pathconf(argv[0],_PC_FILESIZEBITS);
if ( ret == -1)
if ( errno == 0)
printf("_PC_FILESIZEBITS = unlimited\n");
else
printf("_PC_FILESIZEBITS = %s\n", strerror(errno));
else
printf("_PC_FILESIZEBITS = %li\n", ret);
/* ******************** */
errno = 0;
ret = pathconf(argv[0],_PC_LINK_MAX);
if ( ret == -1)
if ( errno == 0)
printf("_PC_LINK_MAX = unlimited\n");
else
printf("_PC_LINK_MAX = %s\n", strerror(errno));
else
printf("_PC_LINK_MAX = %li\n", ret);
/* ******************** */
errno = 0;
ret = pathconf(argv[0],_PC_NAME_MAX);
if ( ret == -1)
if ( errno == 0)
printf("_PC_NAME_MAX = unlimited\n");
else
printf("_PC_NAME_MAX = %s\n", strerror(errno));
else
printf("_PC_NAME_MAX = %li\n", ret);
/* ******************** */
errno = 0;
ret = pathconf(argv[0],_PC_PATH_MAX);
if ( ret == -1)
if ( errno == 0)
printf("_PC_PATH_MAX = unlimited\n");
else
printf("_PC_PATH_MAX = %s\n", strerror(errno));
else
printf("_PC_PATH_MAX = %li\n", ret);
return 0;
}
Zdenek Kotala schrieb:
Regarding to discussion about large segment size of table files a test
pathconf function (see
http://www.opengroup.org/onlinepubs/009695399/functions/pathconf.html).You can see output there:
_PC_FILESIZEBITS - 3rd column
_PC_LINK_MAX - 4th column
_PC_NAME_MAX - 5th column
_PC_PATH_MAX - 6th columnSolaris Nevada ZFS 64 -1 255 1024
UFS 41 32767 255 1024
FAT 33 1 8 1024
NFS 41 32767 255 1024
Solaris 8 UFS 41 32767 255 1024
NFS 40 32767 255 1024
Centos4(2.6.11) EXT3 64 32000 255 4096
XFS 64 2147483647 255 4096
Mac OSX leopard HFS+ 64 32767 255 1024
cygwin 1.5 on NTFS. But 1.7 will a have much larger _PC_PATH_MAX.
_PC_FILESIZEBITS undefined
_PC_LINK_MAX = 8
_PC_NAME_MAX = 260
_PC_PATH_MAX = 257
So this is really bad.
--
Reini Urban
Reini Urban napsal(a):
cygwin 1.5 on NTFS. But 1.7 will a have much larger _PC_PATH_MAX.
_PC_FILESIZEBITS undefined _PC_LINK_MAX = 8 _PC_NAME_MAX = 260 _PC_PATH_MAX = 257
So this is really bad.
Thanks for reporting. It seems not good because postgreSQL assumes that _PC_PATH_MAX
is minimal 1024 for all platforms. If this values is correct then user on cygwin can have a
trouble. Please, could you test postgres with a long PATH?
thanks Zdenek