DPL: add option to force parentFile colocation

[ktf]

No description provided.

[alibuild]

Error while checking build/O2/fullCI_slc9 for 883a847 at 2026-07-06 02:19:

## sw/BUILD/O2-sim-challenge-test-latest/log
./sim-challenge.log[FATAL] TGrid::Connect returned nullptr. May be due to missing alien token
./grp.log[FATAL] TGrid::Connect returned nullptr. May be due to missing alien token


## sw/BUILD/o2checkcode-latest/log
--
========== List of errors found ==========
++ GRERR=0
++ grep -v clang-diagnostic-error error-log.txt
++ grep ' error:'
++ GRERR=1
++ [[ 1 == 0 ]]
++ mkdir -p /sw/INSTALLROOT/15590201305405d349786042af881d9822814aaf/slc9_x86-64/o2checkcode/1.0-local192/etc/modulefiles
++ alibuild-generate-module
+ HAS_BIN=
+ HAS_LIB=
+ HAS_CMAKE=
+ HAS_ROOT=
+ '[' 0 '!=' 0 ']'
+ cat
+ printf 'if ![ is-loaded '\''BASE/1.0'\'' ] {\n module load BASE/1.0\n}'
+ echo 'BUILD_REQUIRES=CMake alibuild-recipe-tools defaults-release'
BUILD_REQUIRES=CMake alibuild-recipe-tools defaults-release
+ FULL_BUILD_REQUIRES='date defaults-release pytorch_cpuinfo FairCMakeModules ninja-fortran CMake googlebenchmark safe_int UUID MPFR Alice-GRID-Utils bz2 double-conversion ninja Python-modules-list O2-customization json-c alibuild-recipe-tools'
+ echo 'FULL_BUILD_REQUIRES=date defaults-release pytorch_cpuinfo FairCMakeModules ninja-fortran CMake googlebenchmark safe_int UUID MPFR Alice-GRID-Utils bz2 double-conversion ninja Python-modules-list O2-customization json-c alibuild-recipe-tools'
FULL_BUILD_REQUIRES=date defaults-release pytorch_cpuinfo FairCMakeModules ninja-fortran CMake googlebenchmark safe_int UUID MPFR Alice-GRID-Utils bz2 double-conversion ninja Python-modules-list O2-customization json-c alibuild-recipe-tools
++ env
++ cut -f1 -d=
++ grep -v '^DEFAULT_'
++ grep -v PKGREVISION
++ grep -v ALIBUILD_RECIPE_TOOLS
++ grep REVISION
++ sed -e s/_REVISION//
+ for x in $(env | cut -f1 -d= | grep -v "^DEFAULT_" | grep -v PKGREVISION | grep -v ALIBUILD_RECIPE_TOOLS | grep REVISION | sed -e 's/_REVISION//')
++ eval 'echo $LIBFFI_REVISION'
+++ echo 10
+ REVISION_VALUE=10
++ eval 'echo $LIBFFI_VERSION'
+++ echo v3.2.1-alice1
+ VERSION_VALUE=v3.2.1-alice1
++ eval 'echo $LIBFFI_ROOT'
+++ echo /sw/slc9_x86-64/libffi/v3.2.1-alice1-10
+ ROOT_PATH_VALUE=/sw/slc9_x86-64/libffi/v3.2.1-alice1-10
+ '[' -z 10 ']'
+ echo 'date defaults-release pytorch_cpuinfo FairCMakeModules ninja-fortran CMake googlebenchmark safe_int UUID MPFR Alice-GRID-Utils bz2 double-conversion ninja Python-modules-list O2-customization json-c alibuild-recipe-tools'
+ tr '[:lower:]' '[:upper:]'
+ tr - _
+ tr ' ' '\n'
[0 more errors; see full log]

Full log here.

[jgrosseo]

@costing I vaguely remember that there are SEs which are counted equivalent because they are close by. Would the check created by @ktf work for this case?

[costing]

I think the exact SE name is checked in the whereis output. Which is ok, the secondary files are supposed to be stored in the same locations as the original ones, not in nearby SEs.

Just that enabling this check for every file will double the load on the catalogue. We already return the list of PFNs at a ::Open() of the original and the secondary files. So if the check could be deferred to when the files are actually about to be used it would not put additional pressure on the databases. In this form it can be used for one time debugging / optimizing step, but it cannot be done regularly or by default.

[jgrosseo]

I agree. This should run as an extra special train and not be enabled for user trains.