Implemented by caching the count at the beginning and showing it instead of the COUNTs
of SubRequest. The latter is available as a -d, --detailed report with all progresses (# of recalls, # of migrations, etc.).

Bug #95189: Time discrepencies between disk servers and name servers can lead to silent data loss on input.

Made the CastorFile.nsOpenTime column NOT NULL and added code to prefill it
at upgrade time. This will be a long operation...

	Add class XrdxCastorClient which handles the async requests and responses

        Move the -D_FILE_OFFSET_BITS=64 flag from the global to local context
        Fix a few errors like unused parameters

Added first draft of ithe stager and the name-server upgrade scripts so they
won't be forgotten.

Fixed in the trunk.

Bug #101372: Incorrect cleaning of old requests leading to wrong final reported statuses of repack -s

Fixed archiving step after a repack abort.

Added the synonym named Vmgr_tape_status_view.

Removed the not yet released or used PL/SQL functions VMGR_GET_TAPE_STATUS and
VMGR_TAPE_IS_RECALLABLE.  They have effectively been replaced by the addition
of the VMGR_TAPE_STATUS_VIEW database view.

Added the VMGR_TAPE_IS_RECALLABLE() function to the VMGR database schema
together with its helper function VMGR_GET_TAPE_STATUS().

Done. Added also the system wide totals as several recall groups might refer to the same tapes. With 3 recall groups we could have 5 + 5 +5 tape being recalled being in fact 3 times the same 5 tapes.

and fall back to the OS uid if not

The sqlplus program interpretts the character as a variable and therefore asks
the user for a value for the variable.

Fixed in the trunk.

Bug #95189: Time discrepencies between disk servers and name servers can lead to silent data loss on input

Fixed a bug created by the commit of revision 25079.  NOT NULL of the stagertime column definition should have been written after the DEFAULT clause.

Fixed in the trunk.

Reverse merging my last commit to make STRINGSTORE return a const char *.  The necessary changes to make the xrootd plugin compile on slc6 are much broader and should be tackled by the responsible developer.

Commited the new xroot-plugin Imakefile from Elvin into the trunk.  The new Imakefile keeps xroot specific compilation constants local to the xroot directory and the associated source files have been renamed from *.cc to *.cpp in order to be compilant with the CASTOR coding and compilation conventions.

With great pain I am effectively reverse merging the commits of Victor with revision numbers 24621 and 24632.  These commits were Victor's adding of support for xroot to the tape servers.  The code worked but we have decided that the xroot API used should be changed.

Second attempt at replacing the incorrect linking with the XrdPosixPreload library with the correct linking to the XrdPosix library

Reverse merging my last commit.  I committed too many files.  May have something to do with the fact that somebody is drilling into the concrete above my head with a very large and annoying drill.

Attempting to replace the incorrect linking with the XrdPosixPreload library with the correct XrdPosix library

  #define _LARGEFILE_SOURCE
  #define _LARGEFILE64_SOURCE
  #define _FILE_OFFSET_BITS 64

Attempting to remove the -D_LARGEFILE64_SOURCE from Imakefile of rtcopy.  This flag was added in such a way that it gets propagated to all of the CPP builds.  If the flag is needed globalling then it should be add to config/Imakefile.tmpl.  If it is needed locally then the rtcopy Imakefile should be modified accordingly

Fixed in the trunk.

The combination of -Wall and -Wextra causes any ununsed variables including parameters to create a warning.  In this commit I have added the -Wno-unused-parameter option to not generate warnings for parameters.  Many thanks to Lukasz for telling us about this very fine grained option.

Bug #95189: Time discrepencies between disk servers and name servers can lead to silent data loss on input

- Properly use cns_file_metadata.stagertime as a single time source for stager open/close operations.
This included modifying a server-side data structure, Cns_file_metadata, and the semantics of
openx() and closex() along with their wire protocol: thus after this change only a new stager can
talk to a new nameserver (i.e. no backwards- nor forwards-compatibility provided).
- Introduced a CastorConfig table in the db with one configuration parameter, openmode, to decide
whether to use the old 'Compatibility' mode as before or the 'New' mode as described above.
- On the way, proper filling of cns_seg_metadata's creationtime, lastmodificationtime and gid fields has
been implemented for logging/statistical purposes.

checking with Elvin that it is not used.

- Cns_setsegattrs
- Cns_replacetapecopy
- Cns_replaceormovetapecopy
This is now deadwood as we moved the functionality to PL/SQL in 2.1.13.