-
Notifications
You must be signed in to change notification settings - Fork 1.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
hadd issue when using parallelization together with indirect file #14910
Comments
Feel free to try this patch: #14913 |
Hi @ferdymercury , thank you for your prompt answer. Maybe I'm doing something wrong, but I wasn't able to get the target file. The patch seems to correctly trigger the parallelization execution:
but then I get:
Were you able to properly merge root files? I compiled the hadd.cxx macro with the following command: |
wrt compression level, did you check the flags to set the compression? |
Unfortunately I had to comment out |
If you do not specify anything, you will get:
Here the full help:
No, I also get a "failed" error. |
Hi @ferdymercury , I think I have found what is causing the error. I believe that here the |
I just edited the patch, try again ;) |
It seems to work properly now, thanks! For which ROOT versions will this fix be available eventually? I was wondering if this modification could be pushed to the LCG release in |
I guess that's a question for @dpiparo or @vepadulano |
As soon as the PR with the patch is merged in master, you will find it in the LCG release of the following day (i.e. at |
Fixes root-project#14910 [skip-ci] document behavior of indirectFileCount
Fixes root-project#14910 [skip-ci] document behavior of indirectFileCount
…14913) Fixes #14910 The full list of input files to merge is now created beforehand by accumulating both the direct and indirect ones (e.g. those found listed in file(s) signaled by the `@` syntax at the command line). Afterwards, the list is used either in the sequential or parallel merge depending on user input. --------- Co-authored-by: Philippe Canal <pcanal@fnal.gov> Co-authored-by: Vincenzo Eduardo Padulano <vincenzo.eduardo.padulano@cern.ch>
…oot-project#14913) Fixes root-project#14910 The full list of input files to merge is now created beforehand by accumulating both the direct and indirect ones (e.g. those found listed in file(s) signaled by the `@` syntax at the command line). Afterwards, the list is used either in the sequential or parallel merge depending on user input. --------- Co-authored-by: Philippe Canal <pcanal@fnal.gov> Co-authored-by: Vincenzo Eduardo Padulano <vincenzo.eduardo.padulano@cern.ch>
…oot-project#14913) Fixes root-project#14910 The full list of input files to merge is now created beforehand by accumulating both the direct and indirect ones (e.g. those found listed in file(s) signaled by the `@` syntax at the command line). Afterwards, the list is used either in the sequential or parallel merge depending on user input. --------- Co-authored-by: Philippe Canal <pcanal@fnal.gov> Co-authored-by: Vincenzo Eduardo Padulano <vincenzo.eduardo.padulano@cern.ch>
…14913) Fixes #14910 The full list of input files to merge is now created beforehand by accumulating both the direct and indirect ones (e.g. those found listed in file(s) signaled by the `@` syntax at the command line). Afterwards, the list is used either in the sequential or parallel merge depending on user input. --------- Co-authored-by: Philippe Canal <pcanal@fnal.gov> Co-authored-by: Vincenzo Eduardo Padulano <vincenzo.eduardo.padulano@cern.ch>
…oot-project#14913) Fixes root-project#14910 The full list of input files to merge is now created beforehand by accumulating both the direct and indirect ones (e.g. those found listed in file(s) signaled by the `@` syntax at the command line). Afterwards, the list is used either in the sequential or parallel merge depending on user input. --------- Co-authored-by: Philippe Canal <pcanal@fnal.gov> Co-authored-by: Vincenzo Eduardo Padulano <vincenzo.eduardo.padulano@cern.ch>
…oot-project#14913) Fixes root-project#14910 The full list of input files to merge is now created beforehand by accumulating both the direct and indirect ones (e.g. those found listed in file(s) signaled by the `@` syntax at the command line). Afterwards, the list is used either in the sequential or parallel merge depending on user input. --------- Co-authored-by: Philippe Canal <pcanal@fnal.gov> Co-authored-by: Vincenzo Eduardo Padulano <vincenzo.eduardo.padulano@cern.ch>
Check duplicate issues.
Description
Dear ROOT experts,
It seems that when running the
hadd
command, the code parallelization is not triggered (e.g.-j 10
) if target files are stored in a .txt file ("indirect file")Reproducer
Input files and .txt file are stored here, I am sorry but I wasn't able to upload root files to git. The issue can be reproduced by running:
hadd -j 10 mkShapes__RDF_2018_v9_emu_ttHMVA.root @doHadd.txt
The initial output says:
Parallelizing with 10 processes
hadd Target file: mkShapes__RDF_2018_v9_emu_ttHMVA.root
hadd compression setting for all output: 1
Each process should handle at least 3 files for efficiency. Setting the number of processes to: 1
The last line is not expected when going to parallel mode, it looks like the script was expecting more arguments from the command line, but target files are stored in the doHadd.txt file
ROOT version
Installation method
/cvmfs/sft.cern.ch/lcg/views/LCG_103/x86_64-centos9-gcc11-opt/bin/root
Operating system
Linux
Additional context
No response
The text was updated successfully, but these errors were encountered: