diff --git a/docs/distributions/c8.md b/docs/distributions/c8.md
deleted file mode 100644
index 77c4d3adeb19cde10a6e902aa57d05bc171de4ff..0000000000000000000000000000000000000000
--- a/docs/distributions/c8.md
+++ /dev/null
@@ -1,190 +0,0 @@
-# CentOS 8
-
-## Practical info
-
-!!! danger ""
-    Upstream support : until 2029-5-31
-
-## Upstream updates
-
-Upstream updates are handled by a [Nomad job](https://gitlab.cern.ch/linuxsupport/cronjobs/centos8_snapshots/).
-
-The job runs automatically every day at 8am. It creates daily snapshot directories
-in `/mnt/data1/dist/cern/centos/8-snapshots/` and maintains the symlinks from `8`,
-`8-testing`, etc. to the appropriate daily snapshot. The `-testing` symlinks
-point to that day's snapshot (see note), and the production (ie. non -testing)
-ones point to the previous Wednesday's snapshot. Production symlinks are changed on
-Wednesdays and ensure packages spend at least 7 days and at most 13 days in testing.
-
-!!! note ""
-    Symlinks are not modified on weekends or between December 16th and January 6th.
-    Snapshots will still be created, though.
-
-
-### Stopping modifications to symlinks
-
-Changes to symlinks can be manually stopped by creating `.freeze.*` files in
-`/mnt/data1/dist/cern/centos/`. You can create the file `.freeze.8all` to stop
-updating all symlinks, or create specific ones for each symlink (ie. `.freeze.8`,
-`.freeze.8.0.1905-testing`, etc.).
-
-A `.freeze.8all` may be created automatically in certain conditions, such as a
-validation failure, detection of a "dangerous" RPM (see below), etc. When this
-happens, an email will be sent to lxsoft-admin explaining the cause. You will
-have to take manual action to resolve the situation and remove the `.freeze.8all`
-so the process can continue.
-
-### *-release packages: the "dangerous" RPMs
-
-The ["dangerous" RPMs](https://gitlab.cern.ch/linuxsupport/cronjobs/centos8_snapshots/-/blob/master/centos8_snapshots/common.sh#L27)
-are generally `*-release` packages. These packages include repo files that point
-to upstream repos, so we must rebuild them to point to our repos. This is
-not done for all `*-release` RPMs, but only for a subset of them for which we
-create snapshots.
-
-We store our new packages on <https://gitlab.cern.ch/linuxsupport/rpms/releases>. Add them
-there if at some point you need to add a new rebuilt package.
-
-As a general rule, when creating a new one, you will need to change the repo file(s) to use
-the corresponding URL from CERN and add the GPG keys. You will also need to add a high epoch
-to make sure CERN's RPM has priority. Use `centos-release` as an example. The source code of
-the upstream packages can be found in the [CentOS git repositories](https://git.centos.org/),
-in the `c8` branch.
-
-Once the new version of a release package is built and tagged as stable, you can
-delete the `.freeze.8all` file and rerun the Nomad job. This time, it will find
-your new RPM in the CERN repo and continue the process.
-
-### Seeing differences between snapshots
-
-You can see all the packages that changed between two dates using the `snapshotdiff`
-command, which is provided by [lxdist-build](https://gitlab.cern.ch/linuxsupport/lxdist-build/-/tree/master/bin):
-
-```bash
-[build@lxsoftadm01] snapshotdiff -d 8 -s 20200701 -e 20200709
-Calculating diff between 20200701 and 20200709...
-
-  apache-sshd                                        2.5.1-1.el8
-  apache-sshd-javadoc                                2.5.1-1.el8
-  bind                                               9.11.13-5.el8_2
-...
-  python-novaclient-doc                              15.1.1-1.el8
-  python-octaviaclient-doc                           1.10.1-1.el8
-  python3-bind                                       9.11.13-5.el8_2
-
-Changes listed: 75
-Total number of changes: 176
-```
-
-### Manually updating symlinks
-
-You can also manually point symlinks to a more up-to-date snapshot. The symlink
-will not be moved backwards in time, but it will move forward again as necessary.
-
-For instance, on Wednesday the 8th, the production symlinks would point to Wednesday
-the 1st. On Thursday the 9th an emergency security patch is released and you manually
-modify the production symlinks to point to that Thurday's snapshot. The links will
-continue pointing to that day even on Wednesday the 15th, where they should
-been moved to the 8th but won't be because they're never moved backwards in time.
-On Wednesday the 22nd they would be moved to Wednesday the 15th as usual.
-
-Before doing this, it is a good idea to check what packages have changed
-between the different snapshots using the `snapshotdiff` command.
-
-The recommended method of updating the links is to ssh into an lxsoftadm machine
-and run:
-
-```bash
-# cd /mnt/data1/dist/cern/centos
-# echo "20200709" > .forcemove.8
-```
-
-Once that file is created, rerun the [snapshotting cron job in Nomad](https://lxsoftadm.cern.ch:4646/ui/jobs/prod_centos8_snapshots). It will see
-this file, set the production symlinks to that date, and then delete the `.forcemove.8`
-file. The job will also send the usual emails and update the web site.
-
-
-## 'Minor' (8.x -> 8.x+1) version release process
-
-The daily update process will do most of the work when a new CentOS 8 minor release
-is published, but there are still some things that will have to be done by hand.
-
-### On the day the release is published by upstream
-
-By the time CentOS announces the new release, our local mirror will already have all
-the new RPMs. We can already prepare things for a smooth rollout the following day:
-
-* Create a new VM with the current CentOS 8 release. We will use it for testing the upgrade.
-    ```bash
-    [aiadm52] ~ > eval `ai-rc "IT Linux Support"`
-    OpenStack environment variables set for Project = 'IT Linux Support'
-    [aiadm52] ~ > openstack image list --property os_distro=C --property os_distro_major=8
-    +--------------------------------------+--------------------------+--------+
-    | ID                                   | Name                     | Status |
-    +--------------------------------------+--------------------------+--------+
-    | 708ba85e-dc9b-49c9-ad73-d645aa87dd67 | C8 - x86_64 [2020-06-01] | active |
-    +--------------------------------------+--------------------------+--------+
-    $ openstack server create --image "C8 - x86_64 [2020-06-01]" --flavor m2.small --property cern-waitdns=false --key-name my-key test-c83
-    ...
-    ```
-* Once that VM is spawned, connect to it, point it to the current testing repos and update it:
-    ```bash
-    [aiadm52] ~ > ssh root@137.138.54.59
-    Enter passphrase for key '/afs/cern.ch/user/a/airibarr/.ssh/id_rsa':
-    X11 forwarding request failed on channel 0
-    [root@test-c83 ~]# cat /etc/dnf/vars/cernbase
-    8
-    [root@test-c83 ~]# echo ".8-testing" > /etc/dnf/vars/cernbase
-    [root@test-c83 ~]# dnf clean all && dnf distro-sync -y
-    0 files removed
-    CentOS-8 - CERN                       388 kB/s |  43 kB     00:00
-    CentOS-8 - AppStream                   18 MB/s | 5.8 MB     00:00
-    CentOS-8 - Base                        20 MB/s | 2.2 MB     00:00
-    ...
-    ```
-* As part of the new CentOS RPMS, there will be a new `centos-release` which [needs to be rebuilt](#-release-packages-the-dangerous-rpms). There may be other `*-release` RPMs to rebuild as well. When you build the packages through Koji, they will automatically be tagged to `cern8-testing`.
-* There will also be a new `kernel`, so a new `kmod-openafs` will need to be built and tagged to `openafs8-testing`.
-* Once that is done, you can go back to your VM and test that the upgrade works as expected. For this, we will use the very latest upstream repos:
-    ```bash
-    [root@test-c83 ~]# echo ".8-upstream" > /etc/dnf/vars/cernbase
-    [root@test-c83 ~]# dnf clean all && dnf distro-sync -y
-    CentOS-8 - CERN                       135 kB/s |  32 kB     00:00
-    CentOS-8 - AppStream                   13 MB/s | 5.8 MB     00:00
-    CentOS-8 - Base                       6.7 MB/s | 2.2 MB     00:00
-    ...
-    ```
-  Note that if you're late in doing this and today's snapshot already contains the latest release, you should use `.8-latest` instead of `.8-upstream`.
-* New AIMS images need to be published. Grab the images from the new release's
-  `/BaseOS/x86_64/kickstart/images/pxeboot/` directory.
-
-    ```bash
-    $ wget http://linuxsoft.cern.ch/cern/centos/8.4/BaseOS/x86_64/kickstart/images/pxeboot/{vmlinuz,initrd.img}
-    ...
-    $ aims2 addimg --testserver --name C84_X86_64_TEST --arch x86_64 --vmlinuz vmlinuz --initrd initrd.img --description "CENTOS 8.4 X86_64 TEST" --uefi
-    $ aims2 addimg --name C84_X86_64_TEST --arch x86_64 --vmlinuz vmlinuz --initrd initrd.img --description "CENTOS 8.4 X86_64 TEST" --uefi
-    ```
-* You may want to do the same with aarch64, but you will not be able to test is as of today due to the lack of HW.
-
-* [Test](/aims2/addinganewimage#testing-a-new-image) the new images
-
-
-### On the following day
-
-If you did all those things yesterday, today you can congratulate yourself and
-go home early. There are just a few more things to do:
-
-* Check that nothing stopped the snapshot process, that the website reflects the
-  [latest updates](https://linux.web.cern.ch/updates/c8/test/latest_updates/) and that the automatic email to `linux-announce-test@cern.ch` was sent. The new release should now be `8-testing`.
-* Send an email to `linux-announce@cern.ch` and `linux-announce-test@cern.ch`
-  announcing the new version and the date that it will become production. Include a link to the release notes.
-* Add the email you just sent to the [news section](https://gitlab.cern.ch/linuxsupport/websites/linux.cern.ch/-/blob/master/docs/index.md) of the website.
-* Let the Config team know that they need to prepare the provisioning templates and rename the OS in Foreman.
-
-### On the day the new release becomes production
-
-Today the new CentOS release becomes production. Just a few more things to do:
-
-* Reupload the AIMS images, but this time without the TEST suffix.
-* Reupload the latest release as the LATEST image.
-* Add entries for previous release to [AIMS hostgroup Hiera data](https://gitlab.cern.ch/ai/it-puppet-hostgroup-aims/-/blob/qa/data/hostgroup/aims.yaml).
-* Remove the previous TEST images.
diff --git a/docs/distributions/cs8.md b/docs/distributions/cs8.md
new file mode 100644
index 0000000000000000000000000000000000000000..c036b80bbafe077befca541749a92097eaefad02
--- /dev/null
+++ b/docs/distributions/cs8.md
@@ -0,0 +1,104 @@
+# CentOS Stream 8
+
+## Practical info
+
+!!! danger ""
+    Upstream support : until 2029-5-31
+
+## Upstream updates
+
+Upstream updates are handled by a [Nomad job](https://gitlab.cern.ch/linuxsupport/cronjobs/stream8_snapshots/).
+
+The job runs automatically every day at 8am. It creates daily snapshot directories
+in `/mnt/data1/dist/cern/centos/s8-snapshots/` and maintains the symlinks from `s8`,
+`s8-testing`, etc. to the appropriate daily snapshot. The `-testing` symlinks
+point to that day's snapshot (see note), and the production (ie. non -testing)
+ones point to the previous Wednesday's snapshot. Production symlinks are changed on
+Wednesdays and ensure packages spend at least 7 days and at most 13 days in testing.
+
+!!! note ""
+    Symlinks are not modified on weekends or between December 16th and January 6th.
+    Snapshots will still be created, though.
+
+
+### Stopping modifications to symlinks
+
+Changes to symlinks can be manually stopped by creating `.freeze.*` files in
+`/mnt/data1/dist/cern/centos/`. You can create the file `.freeze.s8all` to stop
+updating all symlinks, or create specific ones for each symlink (ie. `.freeze.s8`,
+`.freeze.s8.0.1905-testing`, etc.).
+
+A `.freeze.s8all` may be created automatically in certain conditions, such as a
+validation failure, detection of a "dangerous" RPM (see below), etc. When this
+happens, an email will be sent to lxsoft-admin explaining the cause. You will
+have to take manual action to resolve the situation and remove the `.freeze.s8all`
+so the process can continue.
+
+### *-release packages: the "dangerous" RPMs
+
+The ["dangerous" RPMs](https://gitlab.cern.ch/linuxsupport/cronjobs/centos8_snapshots/-/blob/master/centos8_snapshots/common.sh#L27)
+are generally `*-release` packages. These packages include repo files that point
+to upstream repos, so we must rebuild them to point to our repos. This is
+not done for all `*-release` RPMs, but only for a subset of them for which we
+create snapshots.
+
+We store our new packages on <https://gitlab.cern.ch/linuxsupport/rpms/releases>. Add them
+there if at some point you need to add a new rebuilt package.
+
+As a general rule, when creating a new one, you will need to change the repo file(s) to use
+the corresponding URL from CERN and add the GPG keys. You will also need to add a high epoch
+to make sure CERN's RPM has priority. Use `centos-release` as an example. The source code of
+the upstream packages can be found in the [CentOS git repositories](https://git.centos.org/),
+in the `c8s` branch.
+
+Once the new version of a release package is built and tagged as stable, you can
+delete the `.freeze.s8all` file and rerun the Nomad job. This time, it will find
+your new RPM in the CERN repo and continue the process.
+
+### Seeing differences between snapshots
+
+You can see all the packages that changed between two dates using the `snapshotdiff`
+command, which is provided by [lxdist-build](https://gitlab.cern.ch/linuxsupport/lxdist-build/-/tree/master/bin):
+
+```bash
+[build@lxsoftadm01] snapshotdiff -d s8 -s 20200701 -e 20200709
+Calculating diff between 20200701 and 20200709...
+
+  apache-sshd                                        2.5.1-1.el8
+  apache-sshd-javadoc                                2.5.1-1.el8
+  bind                                               9.11.13-5.el8_2
+...
+  python-novaclient-doc                              15.1.1-1.el8
+  python-octaviaclient-doc                           1.10.1-1.el8
+  python3-bind                                       9.11.13-5.el8_2
+
+Changes listed: 75
+Total number of changes: 176
+```
+
+### Manually updating symlinks
+
+You can also manually point symlinks to a more up-to-date snapshot. The symlink
+will not be moved backwards in time, but it will move forward again as necessary.
+
+For instance, on Wednesday the 8th, the production symlinks would point to Wednesday
+the 1st. On Thursday the 9th an emergency security patch is released and you manually
+modify the production symlinks to point to that Thursday's snapshot. The links will
+continue pointing to that day even on Wednesday the 15th, where they should
+been moved to the 8th but won't be because they're never moved backwards in time.
+On Wednesday the 22nd they would be moved to Wednesday the 15th as usual.
+
+Before doing this, it is a good idea to check what packages have changed
+between the different snapshots using the `snapshotdiff` command.
+
+The recommended method of updating the links is to ssh into an lxsoftadm machine
+and run:
+
+```bash
+# cd /mnt/data1/dist/cern/centos
+# echo "20200709" > .forcemove.s8
+```
+
+Once that file is created, rerun the [snapshotting cron job in Nomad](https://lxsoftadm.cern.ch:4646/ui/jobs/prod_stream8_snapshots). It will see
+this file, set the production symlinks to that date, and then delete the `.forcemove.s8`
+file. The job will also send the usual emails and update the web site.
diff --git a/mkdocs.yml b/mkdocs.yml
index 40bd13a43bc1b285e4b25131395077bb1ebe9322..d1d40464ed57828a496470fed3369966cc704c83 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -24,7 +24,7 @@ nav:
     - 'Service Alarms': tasks/alarms.md
     - 'Distributions':
         - 'CC7': distributions/cc7.md
-        - 'C8': distributions/c8.md
+        - 'CS8': distributions/cs8.md
         - 'Docker images': distributions/docker.md
         - 'OpenStack': distributions/openstack.md
         - 'Architecture': distributions/architecture.md