From ec8ea6f73cbcbe5b3a56cb9e42d3d105b56f2814 Mon Sep 17 00:00:00 2001
From: torri <torri@disroot.org>
Date: Thu, 20 Feb 2025 12:49:33 +0100
Subject: [PATCH 01/20] first info for the cvmfsexec way

---
 docs/getting_access/non_standard.md | 130 ++++++++++++++++++++++++++++
 1 file changed, 130 insertions(+)
 create mode 100644 docs/getting_access/non_standard.md

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
new file mode 100644
index 0000000000..eeff61ed32
--- /dev/null
+++ b/docs/getting_access/non_standard.md
@@ -0,0 +1,130 @@
+# Non-standard ways to use EESSI
+
+In case that an EESSI installation is not possible or desirable, there are a couple of "non-standard" ways to access EESSI.
+
+## Using cvmfsexec
+
+This option relies on the [cvmfsexec package](https://github.com/cvmfs/cvmfsexec) provided by CernVM-FS team. It will allow you to mount cvmfs as an unprivileged user, so no need to have a native cvmfs installation.
+
+### `cvmfsexec_eessi.sh`  wrapper script
+
+
+This wrapper script, `cvmfsexec_eessi.sh`, can be used to run a command in a subshell in which the EESSI CernVM-FS repository (`software.eessi.io`) is mounted via cvmfsexec`.
+
+```bash
+#!/bin/bash
+if [ -d /cvmfs/software.eessi.io ]; then
+    # run command directly, EESSI CernVM-FS repository is already mounted
+    "$@"
+else
+    # run command via in subshell where EESSI CernVM-FS repository is mounted,
+    # via cvmfsexec which is set up in a unique temporary directory
+    orig_workdir=$(pwd)
+    mkdir -p /tmp/$USER
+    tmpdir=$(mktemp -p /tmp/$USER -d)
+    cd $tmpdir
+    git clone https://github.com/cvmfs/cvmfsexec.git > $tmpdir/git_clone.out 2>&1
+    cd cvmfsexec
+    ./makedist default > $tmpdir/cvmfsexec_makedist.out 2>&1
+    cd $orig_workdir
+    $tmpdir/cvmfsexec/cvmfsexec software.eessi.io -- "$@"
+    # cleanup
+    rm -rf $tmpdir
+fi
+
+```
+
+Do make sure that this script is executable:
+
+```bash
+    chmod u+x ~/bin/cvmfsexec_eessi.sh
+```
+
+A simple way to test this script is to use it to inspect the contents of the EESSI repository:
+
+```bash
+    ~/bin/cvmfsexec_eessi.sh ls /cvmfs/software.eessi.io
+```
+
+or to start an interactive shell in which the EESSI repository is mounted:
+
+```bash
+    # Expected output from a machine without a cvmfs installation
+    $ ls /cvmfs
+    ls: cannot access '/cvmfs': No such file or directory
+
+    # Starting interactive shell
+    $ ~/bin/cvmfsexec_eessi.sh /bin/bash -l
+    CernVM-FS: loading Fuse module... done
+    CernVM-FS: mounted cvmfs on /tmp/hvela/tmp.iH6e993Adw/cvmfsexec/dist/cvmfs/cvmfs-config.cern.ch
+    CernVM-FS: loading Fuse module... done
+    CernVM-FS: mounted cvmfs on /tmp/hvela/tmp.iH6e993Adw/cvmfsexec/dist/cvmfs/software.eessi.io
+
+    # Now it's mounted!
+    $ ls /cvmfs/
+    cvmfs-config.cern.ch  software.eessi.io
+
+```
+
+Notice how now that EESSI is mounted, you will need to [set up the environment](https://www.eessi.io/docs/using_eessi/setting_up_environment/) in order to access the software itself.
+
+### `orted`  wrapper script
+
+In order to get multi-node runs of software working without having EESSI available system-wide, we also had to create a small wrapper script for the `orted` command that is used by Open MPI to start processes on remote nodes. This is necessary because `mpirun` launches `orted`, which must be run in an environment in which the EESSI repository is mounted. If not, MPI startup will fail with an error like
+
+```bash
+"error: execve(): orted: No such file or directory".
+```
+
+This wrapper script must be named `orted`, and must be located in a path that is listed in `$PATH`.
+
+We placed it in `~/bin/orted`, and you can add `export PATH=$HOME/bin:$PATH` to your `~/.bashrc` login script.
+
+Contents of ~/bin/orted:
+
+```bash
+#!/bin/bash
+
+# first remove path to this orted wrapper from $PATH, to avoid infinite loop
+orted_wrapper_dir=$(dirname $0)
+export PATH=$(echo $PATH | tr ':' '\n' | grep -v $orted_wrapper_dir | tr '\n' ':')
+
+~/bin/cvmfsexec_eessi.sh orted "$@"
+```
+
+Again, do make sure that also this `orted` wrapper script is executable:
+
+```bash
+chmod u+x ~/bin/orted
+```
+
+If not, you will likely run into an error that starts with:
+
+
+```bash
+An ORTE daemon has unexpectedly failed after launch ...
+```
+
+### Slurm job script
+
+The `cvmfsexec_eessi.sh` can be used insided a Slurm job script inside your HPC system to initialize the EESSI environment in a subshell which the EESSI CernVM-FS repository is mounted. Thus, you will be able to load any module from EESSI you might need.
+
+Example job script:
+
+```bash
+#!/bin/bash
+#SBATCH --ntasks=4
+#SBATCH --ntasks-per-node=2
+#SBATCH --cpus-per-task=1
+#SBATCH --time=5:0:0
+#SBATCH --export=None
+#SBATCH --mem=30000M
+~/bin/cvmfsexec_eessi.sh << EOF
+source /cvmfs/software.eessi.io/versions/2023.06/init/bash
+module load TensorFlow/2.13.0-foss-2023a
+export SLURM_EXPORT_ENV=HOME,PATH,LD_LIBRARY_PATH,PYTHONPATH
+mpirun -np 4 python test.py
+EOF
+
+```
+You can see the original blog post on how they used this solution in Deucalion [here](https://www.eessi.io/docs/blog/2024/06/28/espresso-portable-test-run-eurohpc/#running-espresso-on-deucalion-via-eessi-cvmfsexec). 

From 895b7998f4a4a3771cadfa8fc5fc165e2d10112f Mon Sep 17 00:00:00 2001
From: torri <torri@disroot.org>
Date: Thu, 20 Feb 2025 12:53:19 +0100
Subject: [PATCH 02/20] now the index

---
 mkdocs.yml | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mkdocs.yml b/mkdocs.yml
index 9a8c47e21c..a3aad3a2b4 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -33,6 +33,7 @@ nav:
     - Is EESSI already installed?: getting_access/is_eessi_accessible.md
     - Native: getting_access/native_installation.md
     - Container: getting_access/eessi_container.md
+    - Non-standard ways: getting_access/non_standard.md
     - Windows and macOS:
       - Windows with WSL: getting_access/eessi_wsl.md
       - macOS with Lima: getting_access/eessi_limactl.md

From f969e85cfff66d8447e66339c81ef0ee848bf671 Mon Sep 17 00:00:00 2001
From: torri <torri@disroot.org>
Date: Thu, 20 Feb 2025 13:05:27 +0100
Subject: [PATCH 03/20] typo fix

---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index eeff61ed32..d724916218 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -107,7 +107,7 @@ An ORTE daemon has unexpectedly failed after launch ...
 
 ### Slurm job script
 
-The `cvmfsexec_eessi.sh` can be used insided a Slurm job script inside your HPC system to initialize the EESSI environment in a subshell which the EESSI CernVM-FS repository is mounted. Thus, you will be able to load any module from EESSI you might need.
+The `cvmfsexec_eessi.sh` can be used inside a Slurm job script inside your HPC system to initialize the EESSI environment in a subshell which the EESSI CernVM-FS repository is mounted. Thus, you will be able to load any module from EESSI you might need.
 
 Example job script:
 

From 425ccc3017c99d99a31ee9f389b65ba4bea0bdd6 Mon Sep 17 00:00:00 2001
From: torri <torri@disroot.org>
Date: Thu, 20 Feb 2025 16:55:08 +0100
Subject: [PATCH 04/20] WIP on the squashfs part

---
 docs/getting_access/non_standard.md | 105 ++++++++++++++++++++++++++++
 1 file changed, 105 insertions(+)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index d724916218..fbe94f22c7 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -128,3 +128,108 @@ EOF
 
 ```
 You can see the original blog post on how they used this solution in Deucalion [here](https://www.eessi.io/docs/blog/2024/06/28/espresso-portable-test-run-eurohpc/#running-espresso-on-deucalion-via-eessi-cvmfsexec). 
+
+## Via `squashfs` +  cvmfs's `shrinkwrap` utility
+
+CernVM-FS provides the Shrinkwrap utility, allowing users to create a portable snapshot of a CVMFS repository. This can be exported and distributed without the need of a CVMFS client or network access.
+
+To create an export of EESSI in user space, you will first need to create the config file `software.eessi.io.config`:
+
+```bash
+CVMFS_REPOSITORIES=software.eessi.io
+CVMFS_REPOSITORY_NAME=software.eessi.io
+CVMFS_CONFIG_REPOSITORY=cvmfs-config.cern.ch
+CVMFS_SERVER_URL='http://aws-eu-west-s1-sync.eessi.science/cvmfs/software.eessi.io'
+CVMFS_HTTP_PROXY=DIRECT # Avoid filling up any local squid's cache
+CVMFS_CACHE_BASE=/tmp/shrinkwrap
+CVMFS_KEYS_DIR=/etc/cvmfs/keys/eessi.io # Need to be provided for shrinkwrap
+CVMFS_SHARED_CACHE=no # Important as libcvmfs does not support shared caches
+CVMFS_USER=cvmfs
+CVMFS_UID_MAP=uid.map
+CVMFS_GID_MAP=gid.map
+
+```
+You will need to create the files `uid.map` and `gid.map` with the respective value you will use preceded by a `*`. P.e., assuming UID 1000, set the two following files:
+
+```bash
+    $ cat uid.map
+    * 1000
+
+    $ cat gid.map
+    * 1000
+```
+In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. Contents are:
+
+```bash
+/versions/2023.06/compat/linux/x86_64/*
+/versions/2023.06/init/*
+/versions/2023.06/scripts/*
+/versions/2023.06/software/linux/x86_64/intel/skylake_avx512/.lmod/*
+/versions/2023.06/software/linux/x86_64/intel/skylake_avx512/modules/*
+/versions/2023.06/software/linux/x86_64/intel/skylake_avx512/software/*
+# Exclude the Gentoo ebuild repo and cache files
+!/versions/2023.06/compat/linux/x86_64/var/db/repos/gentoo
+!/versions/2023.06/compat/linux/x86_64/var/cache
+
+```
+
+Then, execute `cvmfs_shrinkwrap`to create the export:
+
+```bash
+$ cvmfs_shrinkwrap -r software.eessi.io -f software.eessi.io.config -t software.eessi.io.spec --dest-base /tmp/cvmfs -j 4
+
+LibCvmfs version 2.12, revision 31
+(libcvmfs) (manager 'standard') switching proxy from (none) to DIRECT. Reason: set random start proxy from the first proxy group [Current host: http://aws-eu-west-s1-sync.eessi.science/cvmfs/software.eessi.io]
+(libcvmfs) (manager 'external') switching proxy from (none) to DIRECT. Reason: cloned [Current host: http://aws-eu-west-s1-sync.eessi.science/cvmfs/software.eessi.io]
+(libcvmfs) Starting 4 workers
+(libcvmfs) cntByte|0|Byte count of projected repository
+cntDir|1|Number of directories from source repository
+cntFile|0|Number of files from source repository
+cntSymlink|0|Number of symlinks from source repository
+dataBytes|0|Bytes transferred from source to destination
+dataBytesDeduped|0|Number of bytes not copied due to deduplication
+dataFiles|0|Number of data files transferred from source to destination
+dataFilesDeduped|0|Number of files not copied due to deduplication
+entriesDest|1|Number of file system entries processed in the destination
+entriesSrc|1|Number of file system entries processed in the source
+
+....
+
+# This takes a long time
+```
+
+Once completed, the contents will be available in /tmp/cvmfs. You can create an squashfs image from it:
+
+```bash
+    sudo mksquashfs /tmp/cvmfs software.eessi.io.sqsh
+
+```
+
+This squashfs image can be mounted in any container:
+
+```bash
+    apptainer shell -B software.eessi.io.sqsh:/cvmfs:image-src=/ docker://ubuntu
+
+```
+
+Right now its a manual process but there is work in progress towards an automated solution.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+

From fb4bcaa3ba672c9b476bec6ef9051bf3dcdb8b16 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Mon, 24 Feb 2025 10:44:23 +0100
Subject: [PATCH 05/20] Update non_standard.md

clarirfication for cvmfsexec
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index fbe94f22c7..32832e4904 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -4,7 +4,7 @@ In case that an EESSI installation is not possible or desirable, there are a cou
 
 ## Using cvmfsexec
 
-This option relies on the [cvmfsexec package](https://github.com/cvmfs/cvmfsexec) provided by CernVM-FS team. It will allow you to mount cvmfs as an unprivileged user, so no need to have a native cvmfs installation.
+This option only works for RHEL-like systems and relies on the [cvmfsexec package](https://github.com/cvmfs/cvmfsexec) provided by CernVM-FS team. It will allow you to mount cvmfs as an unprivileged user, so no need to have a native cvmfs installation.
 
 ### `cvmfsexec_eessi.sh`  wrapper script
 

From f836832448c4b6053807c24f48b034cfbe9f0df2 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 20 Mar 2025 10:01:39 +0100
Subject: [PATCH 06/20] Update docs/getting_access/non_standard.md

Co-authored-by: ocaisa <alan.ocais@cecam.org>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 32832e4904..5de144fe6e 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -70,7 +70,7 @@ Notice how now that EESSI is mounted, you will need to [set up the environment](
 
 ### `orted`  wrapper script
 
-In order to get multi-node runs of software working without having EESSI available system-wide, we also had to create a small wrapper script for the `orted` command that is used by Open MPI to start processes on remote nodes. This is necessary because `mpirun` launches `orted`, which must be run in an environment in which the EESSI repository is mounted. If not, MPI startup will fail with an error like
+In order to get multi-node runs of software working without having EESSI available system-wide, you also need to create a small wrapper script for the `orted` command that is used by Open MPI to start processes on remote nodes. This is necessary because `mpirun` launches `orted`, which must be run in an environment in which the EESSI repository is mounted. If not, MPI startup may fail with an error like
 
 ```bash
 "error: execve(): orted: No such file or directory".

From 5736bbc7221601da42f856c6720718b0255c9077 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 20 Mar 2025 10:02:01 +0100
Subject: [PATCH 07/20] Update docs/getting_access/non_standard.md

Co-authored-by: ocaisa <alan.ocais@cecam.org>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 5de144fe6e..31f920930d 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -78,7 +78,7 @@ In order to get multi-node runs of software working without having EESSI availab
 
 This wrapper script must be named `orted`, and must be located in a path that is listed in `$PATH`.
 
-We placed it in `~/bin/orted`, and you can add `export PATH=$HOME/bin:$PATH` to your `~/.bashrc` login script.
+For example, it can be placed in `~/bin/orted`, and you can add `export PATH=$HOME/bin:$PATH` to your `~/.bashrc` login script.
 
 Contents of ~/bin/orted:
 

From b3ba9a55828dd6a8121f71abd404dc2603132f03 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 20 Mar 2025 10:02:25 +0100
Subject: [PATCH 08/20] Update docs/getting_access/non_standard.md

Co-authored-by: ocaisa <alan.ocais@cecam.org>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 31f920930d..0c957b9487 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -212,7 +212,7 @@ This squashfs image can be mounted in any container:
 
 ```
 
-Right now its a manual process but there is work in progress towards an automated solution.
+Right now, this may seem like quite a manual process, but there is work in progress towards creating a script to automate this process.
 
 
 

From e5600c9b0187be2b95d0c36906a5936b8e28c357 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 20 Mar 2025 10:02:56 +0100
Subject: [PATCH 09/20] Update docs/getting_access/non_standard.md

Co-authored-by: ocaisa <alan.ocais@cecam.org>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 0c957b9487..867a06a0f0 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -149,7 +149,7 @@ CVMFS_UID_MAP=uid.map
 CVMFS_GID_MAP=gid.map
 
 ```
-You will need to create the files `uid.map` and `gid.map` with the respective value you will use preceded by a `*`. P.e., assuming UID 1000, set the two following files:
+You will need to create the files `uid.map` and `gid.map` with the respective value you will use preceded by a `*`. For example, assuming UID 1000, set the two following files:
 
 ```bash
     $ cat uid.map

From d77b762e42d41e124792b65a4d7abf34604a525e Mon Sep 17 00:00:00 2001
From: torri <torri@disroot.org>
Date: Thu, 20 Mar 2025 15:31:34 +0100
Subject: [PATCH 10/20] first batch of corrections

---
 docs/getting_access/non_standard.md | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 867a06a0f0..8e32e00a12 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -131,7 +131,7 @@ You can see the original blog post on how they used this solution in Deucalion [
 
 ## Via `squashfs` +  cvmfs's `shrinkwrap` utility
 
-CernVM-FS provides the Shrinkwrap utility, allowing users to create a portable snapshot of a CVMFS repository. This can be exported and distributed without the need of a CVMFS client or network access.
+CernVM-FS provides the [Shrinkwrap utility](https://cvmfs.readthedocs.io/en/stable/cpt-shrinkwrap.html), allowing users to create a portable snapshot of a CVMFS repository. This can be exported and distributed without the need of a CVMFS client or network access.
 
 To create an export of EESSI in user space, you will first need to create the config file `software.eessi.io.config`:
 
@@ -149,14 +149,16 @@ CVMFS_UID_MAP=uid.map
 CVMFS_GID_MAP=gid.map
 
 ```
-You will need to create the files `uid.map` and `gid.map` with the respective value you will use preceded by a `*`. For example, assuming UID 1000, set the two following files:
+You will need to create the files `uid.map` and `gid.map` with the respective value you will use preceded by a `*`. You will probably want to use the UID and GID of the current user, so, to set the two files:
 
 ```bash
+    $ echo '*' $(id -u $USER) > uid.map
     $ cat uid.map
-    * 1000
+    * 1001
 
+    $ echo '*' $(id -g $USER) > gid.map
     $ cat gid.map
-    * 1000
+    * 1001
 ```
 In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. Contents are:
 
@@ -195,13 +197,13 @@ entriesSrc|1|Number of file system entries processed in the source
 
 ....
 
-# This takes a long time
+# This takes a long time and space depending on how much you want to export
 ```
 
 Once completed, the contents will be available in /tmp/cvmfs. You can create an squashfs image from it:
 
 ```bash
-    sudo mksquashfs /tmp/cvmfs software.eessi.io.sqsh
+    mksquashfs /tmp/cvmfs software.eessi.io.sqsh
 
 ```
 

From 5a9d473185af58a454341270f799e0f12e76b485 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Fri, 14 Nov 2025 09:34:14 +0100
Subject: [PATCH 11/20] Update docs/getting_access/non_standard.md

Co-authored-by: Caspar van Leeuwen <33718780+casparvl@users.noreply.github.com>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 8e32e00a12..6ac6fae8fe 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -4,7 +4,7 @@ In case that an EESSI installation is not possible or desirable, there are a cou
 
 ## Using cvmfsexec
 
-This option only works for RHEL-like systems and relies on the [cvmfsexec package](https://github.com/cvmfs/cvmfsexec) provided by CernVM-FS team. It will allow you to mount cvmfs as an unprivileged user, so no need to have a native cvmfs installation.
+This option only works for [RHEL-like systems](https://github.com/cvmfs/cvmfsexec?tab=readme-ov-file#supported-operating-systems) and relies on the [cvmfsexec package](https://github.com/cvmfs/cvmfsexec) provided by CernVM-FS team. It will allow you to mount cvmfs as an unprivileged user, so no need to have a native cvmfs installation.
 
 ### `cvmfsexec_eessi.sh`  wrapper script
 

From 4d14e1ad48b4dae8b99b23005340404838787f34 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Fri, 14 Nov 2025 09:34:37 +0100
Subject: [PATCH 12/20] Update docs/getting_access/non_standard.md

Co-authored-by: Caspar van Leeuwen <33718780+casparvl@users.noreply.github.com>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 6ac6fae8fe..b50ee1b538 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -127,7 +127,7 @@ mpirun -np 4 python test.py
 EOF
 
 ```
-You can see the original blog post on how they used this solution in Deucalion [here](https://www.eessi.io/docs/blog/2024/06/28/espresso-portable-test-run-eurohpc/#running-espresso-on-deucalion-via-eessi-cvmfsexec). 
+You can see the original blog post on how this solution was used on the Deucalion EuroHPC system [here](https://www.eessi.io/docs/blog/2024/06/28/espresso-portable-test-run-eurohpc/#running-espresso-on-deucalion-via-eessi-cvmfsexec). 
 
 ## Via `squashfs` +  cvmfs's `shrinkwrap` utility
 

From af1bf6481348833962e7f1d61192ee446a64d4aa Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Fri, 14 Nov 2025 09:45:44 +0100
Subject: [PATCH 13/20] Update docs/getting_access/non_standard.md

Co-authored-by: Caspar van Leeuwen <33718780+casparvl@users.noreply.github.com>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index b50ee1b538..bb366463e7 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -131,7 +131,7 @@ You can see the original blog post on how this solution was used on the Deucalio
 
 ## Via `squashfs` +  cvmfs's `shrinkwrap` utility
 
-CernVM-FS provides the [Shrinkwrap utility](https://cvmfs.readthedocs.io/en/stable/cpt-shrinkwrap.html), allowing users to create a portable snapshot of a CVMFS repository. This can be exported and distributed without the need of a CVMFS client or network access.
+CernVM-FS provides the [Shrinkwrap utility](https://cvmfs.readthedocs.io/en/stable/cpt-shrinkwrap.html), allowing users to create a portable snapshot of a CVMFS repository. This can be exported and distributed without the need of a CVMFS client or network access. A big advantage of `Shrinkwrap` over e.g. simply `rsync`-ing the repository to a local directory `/cvmfs/software.eessi.io` is that `Shrinkwrap` preserves the deduplication that CernVM-FS offers.
 
 To create an export of EESSI in user space, you will first need to create the config file `software.eessi.io.config`:
 

From f9084d9494d9a93153aea60e3a83b7c973eeb0e8 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Fri, 14 Nov 2025 09:46:02 +0100
Subject: [PATCH 14/20] Update docs/getting_access/non_standard.md

Co-authored-by: Caspar van Leeuwen <33718780+casparvl@users.noreply.github.com>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index bb366463e7..c5f85033c0 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -197,7 +197,7 @@ entriesSrc|1|Number of file system entries processed in the source
 
 ....
 
-# This takes a long time and space depending on how much you want to export
+# This takes a long time, memory and storage space depending on how much you want to export
 ```
 
 Once completed, the contents will be available in /tmp/cvmfs. You can create an squashfs image from it:

From 7f218cedff52dd26f07aa9a284aefd734691c333 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Wed, 26 Nov 2025 15:55:06 +0100
Subject: [PATCH 15/20] Apply suggestion from @casparvl

Co-authored-by: Caspar van Leeuwen <33718780+casparvl@users.noreply.github.com>
---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index c5f85033c0..ecee4fd333 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -160,7 +160,7 @@ You will need to create the files `uid.map` and `gid.map` with the respective va
     $ cat gid.map
     * 1001
 ```
-In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. Contents are:
+In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. For example:
 
 ```bash
 /versions/2023.06/compat/linux/x86_64/*

From 8163d0ffde00356a07f5d9e896fae00234e2f8a4 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 27 Nov 2025 14:28:36 +0000
Subject: [PATCH 16/20] added compression options

---
 docs/getting_access/non_standard.md | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index ecee4fd333..c178b88bb6 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -160,7 +160,7 @@ You will need to create the files `uid.map` and `gid.map` with the respective va
     $ cat gid.map
     * 1001
 ```
-In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. For example:
+In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. A basic setup would be:
 
 ```bash
 /versions/2023.06/compat/linux/x86_64/*
@@ -174,6 +174,7 @@ In addition, you need to create a spec file `software.eessi.io.spec` with the fi
 !/versions/2023.06/compat/linux/x86_64/var/cache
 
 ```
+[WIP] Please note that specifying the compatibility layer is a must to ensure a proper functioning ? of the export, as it initializes the environment. 
 
 Then, execute `cvmfs_shrinkwrap`to create the export:
 
@@ -206,6 +207,11 @@ Once completed, the contents will be available in /tmp/cvmfs. You can create an
     mksquashfs /tmp/cvmfs software.eessi.io.sqsh
 
 ```
+Note how `mksquashfs` compresses by default all generated images using `gzip`. If you want a faster approach at compression, you can use the `-comp` option and specify your preferred option among the available ones (you can check them with the `--help` argument). We recommend `zstd` because it's multithreaded and typically achieves better compression:
+
+```
+mksquashfs /tmp/cvmfs software.eessi.io.sqsh -comp zstd  
+```
 
 This squashfs image can be mounted in any container:
 

From 26ab3b7451e4c5bd0a2815720a1f8f2cc646bf9e Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 27 Nov 2025 15:13:32 +0000
Subject: [PATCH 17/20] adding explanation in the spec files

---
 docs/getting_access/non_standard.md | 76 +++++++++++++++++++++++++++--
 1 file changed, 72 insertions(+), 4 deletions(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index c178b88bb6..919063be38 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -160,7 +160,19 @@ You will need to create the files `uid.map` and `gid.map` with the respective va
     $ cat gid.map
     * 1001
 ```
-In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. A basic setup would be:
+In addition, you need to create a spec file `software.eessi.io.spec` with the files you want to include and/or exclude in the shrinkwrap. Please note that specifying the compatibility layer is a must to ensure a proper functioning of the export as it initializes the environment. The most basic setup would be:
+
+```bash
+/versions/2023.06/compat/linux/x86_64/*
+/versions/2023.06/init/*
+/versions/2023.06/scripts/*
+# Exclude the Gentoo ebuild repo and cache files
+!/versions/2023.06/compat/linux/x86_64/var/db/repos/gentoo
+!/versions/2023.06/compat/linux/x86_64/var/cache
+
+``` 
+From here, you can tune this file to your needs, you can change the release version, the architecture and/or micro-architectures, or even if you just need one particular software. For example, if you wanted to have all the software for Intel's AVX512, you would need to export:  
+
 
 ```bash
 /versions/2023.06/compat/linux/x86_64/*
@@ -174,7 +186,55 @@ In addition, you need to create a spec file `software.eessi.io.spec` with the fi
 !/versions/2023.06/compat/linux/x86_64/var/cache
 
 ```
-[WIP] Please note that specifying the compatibility layer is a must to ensure a proper functioning ? of the export, as it initializes the environment. 
+
+If instead you wanted the whole tree for a handful of another micro-architecture: 
+
+```bash
+/versions/2023.06/compat/linux/x86_64/*
+/versions/2023.06/init/*
+/versions/2023.06/scripts/*
+/versions/2023.06/software/linux/aarch64/neoverse_n1/*
+/versions/2023.06/software/linux/aarch64/a64fx/*
+/versions/2023.06/software/linux/aarch64/nvidia/*
+# Exclude the Gentoo ebuild repo and cache files
+!/versions/2023.06/compat/linux/x86_64/var/db/repos/gentoo
+!/versions/2023.06/compat/linux/x86_64/var/cache
+
+```
+But be careful as exporting the whole tree takes huge amounts of memory. 
+
+If you want a cleaner approach, you can store the compat layer in a separate file called `software.eessi.io_compat.spec` with the minimal contents stated before. And in the  `software.eessi.io.spec ` only the software files you are interested in. Per example, if you were only interested in GROMACS, you could use the environment inside an existing EESSI installation to generate the spec file from the loaded modules: 
+
+``` bash
+# Init EESSI and load the target software
+source /cvmfs/software.eessi.io/versions/2023.06/init/lmod/bash
+module load GROMACS
+
+# Create the spec file for software
+ls /cvmfs/software.eessi.io/versions/$EESSI_VERSION/software/linux/$EESSI_SOFTWARE_SUBDIR/modules -d | sed s#/cvmfs/software.eessi.io##g > software.eessi.io.spec
+ls /cvmfs/software.eessi.io/versions/$EESSI_VERSION/software/linux/$EESSI_SOFTWARE_SUBDIR/.lmod -d | sed s#/cvmfs/software.eessi.io##g >> software.eessi.io.spec
+env | grep EBROOT | sed s/=/\ /g | awk '{print $2}' | sed s#/cvmfs/software.eessi.io##g | xargs -i echo "{}/*" >> software.eessi.io.spec
+``
+
+The `software.eessi.io.spec` file will look this way: 
+
+```bash 
+/versions/2023.06/software/linux/x86_64/intel/haswell/modules
+/versions/2023.06/software/linux/x86_64/intel/haswell/.lmod
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/Python-bundle-PyPI/2023.10-GCCcore-13.2.0/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/pybind11/2.11.1-GCCcore-13.2.0/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/gfbf/2023b/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/PMIx/4.2.6-GCCcore-13.2.0/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/OpenSSL/1.1/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/foss/2023b/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/OpenMPI/4.1.6-GCC-13.2.0/*
+[...]
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/GROMACS/2024.4-foss-2023b/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/gompi/2023b/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/GCCcore/13.2.0/*
+/versions/2023.06/software/linux/x86_64/intel/haswell/software/SQLite/3.43.1-GCCcore-13.2.0/*
+
+```
 
 Then, execute `cvmfs_shrinkwrap`to create the export:
 
@@ -201,10 +261,18 @@ entriesSrc|1|Number of file system entries processed in the source
 # This takes a long time, memory and storage space depending on how much you want to export
 ```
 
+If you have done the compat layer and software specs separetly, you will also need to make the export of the compat layer apart: 
+
+```bash
+cvmfs_shrinkwrap -r software.eessi.io -f software.eessi.io.config -t software.eessi.io_compat.spec --dest-base /tmp/compat -j 4
+```
+
 Once completed, the contents will be available in /tmp/cvmfs. You can create an squashfs image from it:
 
 ```bash
-    mksquashfs /tmp/cvmfs software.eessi.io.sqsh
+mksquashfs /tmp/cvmfs software.eessi.io.sqsh
+# And if you squashed the compat layer separetly: 
+mksquashfs /tmp/compat software.eessi.io_compat.sqsh
 
 ```
 Note how `mksquashfs` compresses by default all generated images using `gzip`. If you want a faster approach at compression, you can use the `-comp` option and specify your preferred option among the available ones (you can check them with the `--help` argument). We recommend `zstd` because it's multithreaded and typically achieves better compression:
@@ -216,7 +284,7 @@ mksquashfs /tmp/cvmfs software.eessi.io.sqsh -comp zstd
 This squashfs image can be mounted in any container:
 
 ```bash
-    apptainer shell -B software.eessi.io.sqsh:/cvmfs:image-src=/ docker://ubuntu
+apptainer shell -B software.eessi.io.sqsh:/cvmfs:image-src=/ docker://ubuntu
 
 ```
 

From 74319f487ed6e65fe51a55a504d9b4c19b1e5ae9 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 27 Nov 2025 16:12:51 +0000
Subject: [PATCH 18/20] typo fix

---
 docs/getting_access/non_standard.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 919063be38..efcf7ca69e 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -214,7 +214,7 @@ module load GROMACS
 ls /cvmfs/software.eessi.io/versions/$EESSI_VERSION/software/linux/$EESSI_SOFTWARE_SUBDIR/modules -d | sed s#/cvmfs/software.eessi.io##g > software.eessi.io.spec
 ls /cvmfs/software.eessi.io/versions/$EESSI_VERSION/software/linux/$EESSI_SOFTWARE_SUBDIR/.lmod -d | sed s#/cvmfs/software.eessi.io##g >> software.eessi.io.spec
 env | grep EBROOT | sed s/=/\ /g | awk '{print $2}' | sed s#/cvmfs/software.eessi.io##g | xargs -i echo "{}/*" >> software.eessi.io.spec
-``
+``` 
 
 The `software.eessi.io.spec` file will look this way: 
 

From ffd2bf7715b9e17cad801e1c47a6555180a50159 Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 27 Nov 2025 16:14:14 +0000
Subject: [PATCH 19/20] typo fix part 2

---
 docs/getting_access/non_standard.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index efcf7ca69e..75cc3324f1 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -261,7 +261,7 @@ entriesSrc|1|Number of file system entries processed in the source
 # This takes a long time, memory and storage space depending on how much you want to export
 ```
 
-If you have done the compat layer and software specs separetly, you will also need to make the export of the compat layer apart: 
+If you have done the compat layer and software specs separately, you will also need to make the export of the compat layer apart: 
 
 ```bash
 cvmfs_shrinkwrap -r software.eessi.io -f software.eessi.io.config -t software.eessi.io_compat.spec --dest-base /tmp/compat -j 4
@@ -271,7 +271,7 @@ Once completed, the contents will be available in /tmp/cvmfs. You can create an
 
 ```bash
 mksquashfs /tmp/cvmfs software.eessi.io.sqsh
-# And if you squashed the compat layer separetly: 
+# And if you squashed the compat layer separately: 
 mksquashfs /tmp/compat software.eessi.io_compat.sqsh
 
 ```

From 85c39bc656545f6fa8f6c6a6c0437b040693ee4a Mon Sep 17 00:00:00 2001
From: Helena Vela Beltran <47674829+hvelab@users.noreply.github.com>
Date: Thu, 27 Nov 2025 16:21:41 +0000
Subject: [PATCH 20/20] adding section

---
 docs/getting_access/non_standard.md | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/docs/getting_access/non_standard.md b/docs/getting_access/non_standard.md
index 75cc3324f1..9c5585de15 100644
--- a/docs/getting_access/non_standard.md
+++ b/docs/getting_access/non_standard.md
@@ -133,6 +133,8 @@ You can see the original blog post on how this solution was used on the Deucalio
 
 CernVM-FS provides the [Shrinkwrap utility](https://cvmfs.readthedocs.io/en/stable/cpt-shrinkwrap.html), allowing users to create a portable snapshot of a CVMFS repository. This can be exported and distributed without the need of a CVMFS client or network access. A big advantage of `Shrinkwrap` over e.g. simply `rsync`-ing the repository to a local directory `/cvmfs/software.eessi.io` is that `Shrinkwrap` preserves the deduplication that CernVM-FS offers.
 
+Please note that this utility requires a system that already does have access EESSI, and once exported, you can use that export elsewhere you might need it. If you have access to a system with EESSI installed but no sudo rights, it's perfectly fine as the export can be done in user space. If you don't have access to a system with EESSI, you could do it on your own PC where you can install it by yourself or do it inside the container, but be careful because it requires a huge amount of memory and a long time to create the export. 
+
 To create an export of EESSI in user space, you will first need to create the config file `software.eessi.io.config`:
 
 ```bash