Hi Yelena,
we had running Robinhood on Debian Jessie some time ago...
I built the Debian package for Robinhood from the sources with
help of GIT and Debian Build Tools.
Since we moved to CentOS we do not build any new Robinhood packages for
Debian anymore. But The following instructions did the job.
Hope that helps.
Best regards
Gabriele
**Building a Debian Package for Robinhood from a GIT Repository**
1. Clone the Robinhood GIT repository
git clone https://github.com/cea-hpc/robinhood.git robinhood-code
cd robinhood-code
2. Checkout tag version 3.0
git checkout 3.0
3. Create source TAR ball
git archive 3.0 | gzip > ../robinhood_3.0.orig.tar.gz
cd ..
4. Create an GIT repository
mkdir robinhood_3.0
cd robinhood_3.0
git init
5. Import the GIT repository from the TAR ball
git-import-orig -u 3.0 ../robinhood_3.0.orig.tar.gz
6. Prepare Debian Packaging
dh_make -s -n -e g.ianne...@gsi.de -p robinhood_3.0
7. Edit Debian Changelog File
vi debian/changelog
set version string to (3.0-1~gsi8+1)
set package repository to jessie-testing
8. Edit Debian Source Format File
vi debian/source/format
1.0
9. Edit Debian Control File
vi debian/control
Section: misc
Hompage: https://github.com/cea-hpc/robinhood/wiki
Depends: lustre-dev (>= 2.8)
Description: Robinhood manages contents of large file systems.
Robinhood Policy Engine is a versatile tool to manage contents of
large file systems. It maintains a replicate of filesystem medatada in a
database that can be queried at will. It makes it possible to schedule
mass action on filesystem entries by defining attribute-based policies,
provides fast 'find' and 'du' enhanced clones, gives to administrators
an overall view of filesystem contents through its web UI and command
line tools. It supports any POSIX filesystem and implements advanced
features for Lustre filesystems (list/purge files per OST or pool, read
MDT changelogs...)
10. Create Robinhood Service-Unit File
vi debian/robinhood.service
[Unit]
Description=Robinhood Policy Engine
[Service]
Type=simple
ExecStart=/usr/sbin/robinhood --run=all -r
[Install]
WantedBy=multi-user.target
11. Execute autogen
./autogen.sh
12. Edit Debian Rules File
vi debian/rules
override_dh_auto_configure:
{TAB} dh_auto_configure -- \
{TAB} --with-lustre=/home/iannetti/workspace/lustre/lustre-release/
ALTERNATIVE: Use "--enable-lustre" than "with-lustre=..."
13. Commit Debian Directory to GIT Repository
git add debian
git commit -a -m "Added Debian Directory"
14. Provide Lustre Library File
sudo cp
/home/iannetti/workspace/lustre/lustre-release/lustre/utils/liblustreapi.so
/usr/lib/
ALTERNATIVE: Install the Lustre dev package to prevent copying shared
library.
15. Build the Debian Package
git-buildpackage -uc -us --git-ignore-branch
--git-upstream-tag="%(version)s"
On 6/10/20 3:47 PM, Yelena Gelzayd wrote:
Hi,
I'm looking to test out Robinhood to analyze large NFS-mounted filesystems.
We only run Debian. The instructions to install Robinhood only talk
about compiling and building an rpm.
My questions are:
- has anyone installed Robinhood on Debian?
- is it just a matter of using something like animal to convert an rpm
to deb package? (i would not want to install an rpm on Debian and create
a non-standard setup)
- has anyone used Robinhood to scan NFS-mounted filesystems? Petabytes
in size, billions of files.
- how active is Robinhood development?
Thank you in advance,
Yelena
_______________________________________________
robinhood-support mailing list
robinhood-support@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/robinhood-support
_______________________________________________
robinhood-support mailing list
robinhood-support@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/robinhood-support