. _______________________________________________
...
Has any tested openafs on sun4x_56 or sun5x_57?
_______________________________________________
> sun4x_56 <
After the readdir patch, an OpenAFS-1.0.2 client is not hanging any more
in 'ls' for a big AFS directory on this machine type :
SunOS 5.6 Generic_105181-23 sun4m sparc SUNW,SPARCstation-5
> sun4x_58 <
This type of OpenAFS-1.0.2 VL + file server previously was hanging with 'ls'
in a big AFS directory:
SunOS 5.8 Generic_108528-04 sun4m sparc SUNW,SPARCstation-20
, allocating all swap space (took some time). Finally returned something
like 'resource not available'.
The server part did already function well.
After a crash, its /vicepa did not mount. So I compiled the vfsck for
sun4x_58 which did recover it well.
Another compilation with your readdir patch applied just finished. Now doing an
'ls /afs' , I can see 160+ entries on above machine. Fine.
So, sun4x_57 presumably is no problem any more.
Thanks a lot to you for finding out the reason!
Best regards,
Rainer Laatsch [EMAIL PROTECTED]
PS. We use the following text of your patch for readdir (no #if 0):
*** openafs-1.0.2.old/src/afs/VNOPS/afs_vnop_readdir.c Thu Feb 1 16:10:03 2001
--- openafs-1.0.2/src/afs/VNOPS/afs_vnop_readdir.c Thu Feb 1 16:09:19 2001
***************
*** 641,651 ****
}
#else /* AFS_HPUX_ENV */
code = afs_readdir_move(ode, avc, auio, o_slen,
- #if defined(AFS_SUN5_ENV)
- len, origOffset);
- #else
auio->afsio_resid, origOffset);
- #endif
#endif /* AFS_HPUX_ENV */
/* this next line used to be AFSVFS40 or AIX 3.1, but is
* really generic */
--- 641,647 ----
_______________________________________________
OpenAFS-info mailing list
[EMAIL PROTECTED]
https://lists.openafs.org/mailman/listinfo.cgi/openafs-info
