Support Valkey 7 in CentOS Stream 10

[phracek]

This pull request enables building and testing valkey-container in CentOS Stream 10

[github-actions]

Pull Request validation

Failed

🔴 Failed or pending statuses - Testing Farm - CentOS Stream 10 - 7[pending],Testing Farm - Fedora - 7[pending]
🔴 Review - Missing review from a member

[phracek]

Difference between Dockerfile.fedora and redis dockerfile.fedora in version 7

$ diff -u Dockerfile.fedora ../../redis-container/7/Dockerfile.fedora
--- Dockerfile.fedora	2024-08-16 09:53:12
+++ ../../redis-container/7/Dockerfile.fedora	2024-08-15 13:59:00
@@ -1,70 +1,62 @@
-FROM quay.io/fedora/s2i-core:40
+FROM quay.io/fedora/s2i-core:38

-# Valkey image based on Software Collections packages
+# Redis image based on Software Collections packages
 #
 # Volumes:
-#  * /var/lib/valkey/data - Datastore for Valkey
+#  * /var/lib/redis/data - Datastore for Redis
 # Environment:
-#  * $VALKEY_PASSWORD - Database password
+#  * $REDIS_PASSWORD - Database password

-ENV VALKEY_VERSION=7 \
-    HOME=/var/lib/valkey \
-    NAME=valkey
+ENV REDIS_VERSION=7 \
+    HOME=/var/lib/redis

-ENV SUMMARY="Valkey in-memory data structure store, used as database, cache and message broker" \
-    DESCRIPTION="Valkey $VALKEY_VERSION available as container, is an advanced key-value store. \
+ENV SUMMARY="Redis in-memory data structure store, used as database, cache and message broker" \
+    DESCRIPTION="Redis $REDIS_VERSION available as container, is an advanced key-value store. \
 It is often referred to as a data structure server since keys can contain strings, hashes, lists, \
 sets and sorted sets. You can run atomic operations on these types, like appending to a string; \
 incrementing the value in a hash; pushing to a list; computing set intersection, union and difference; \
 or getting the member with highest ranking in a sorted set. In order to achieve its outstanding \
-performance, Valkey works with an in-memory dataset. Depending on your use case, you can persist \
+performance, Redis works with an in-memory dataset. Depending on your use case, you can persist \
 it either by dumping the dataset to disk every once in a while, or by appending each command to a log."

 LABEL summary="$SUMMARY" \
       description="$DESCRIPTION" \
       io.k8s.description="$DESCRIPTION" \
-      io.k8s.display-name="Valkey $VALKEY_VERSION" \
-      io.openshift.expose-services="6379:valkey" \
-      io.openshift.tags="database,valkey,valkey$VALKEY_VERSION,valkey-$VALKEY_VERSION" \
+      io.k8s.display-name="Redis $REDIS_VERSION" \
+      io.openshift.expose-services="6379:redis" \
+      io.openshift.tags="database,redis,redis$REDIS_VERSION,redis-$REDIS_VERSION" \
       com.redhat.component="$NAME" \
-      name="fedora/$NAME-$VALKEY_VERSION" \
-      version="$VALKEY_VERSION" \
-      usage="podman run -d --name valkey_database -p 6379:6379 quay.io/fedora/$NAME-$VALKEY_VERSION" \
+      name="fedora/$NAME-$REDIS_VERSION" \
+      version="$REDIS_VERSION" \
+      usage="podman run -d --name redis_database -p 6379:6379 quay.io/fedora/$NAME-$REDIS_VERSION" \
       maintainer="SoftwareCollections.org <[email protected]>"

 EXPOSE 6379

-# Get prefix path and path to scripts rather than hard-code them in scripts
-ENV CONTAINER_SCRIPTS_PATH=/usr/share/container-scripts/valkey \
-    VALKEY_PREFIX=/usr \
-    VALKEY_CONF=/etc/valkey/valkey.conf \
-    VALKEY_SOCK=/run/valkey/valkey.sock \
-    VALKEY_LIB=/var/lib/valkey \
-    VALKEY_RUN=/run/valkey
-
-# Create user for valkey that has known UID
+# Create user for redis that has known UID
 # We need to do this before installing the RPMs which would create user with random UID
 # The UID is the one used by the default user from the parent layer (1001),
 # and since the user exists already, do not create a new one, but only rename
 # the existing
-# This image must forever use UID 1001 for valkey user so our volumes are
+# This image must forever use UID 1001 for redis user so our volumes are
 # safe in the future. This should *never* change, the last test is there
 # to make sure of that.
-RUN getent group  valkey &> /dev/null || groupadd -r valkey &> /dev/null && \
-    usermod -l valkey -aG valkey -c 'Server' default &> /dev/null && \
+RUN getent group  redis &> /dev/null || groupadd -r redis &> /dev/null && \
+    usermod -l redis -aG redis -c 'Redis Server' default &> /dev/null && \
 # Install gettext for envsubst command
-    dnf install -y dnf-utils gettext policycoreutils && \
-    INSTALL_PKGS="valkey" && \
+    dnf install -y yum-utils gettext policycoreutils && \
+    INSTALL_PKGS="redis" && \
     dnf install -y --setopt=tsflags=nodocs --nogpgcheck $INSTALL_PKGS && \
     rpm -V $INSTALL_PKGS && \
     dnf -y clean all --enablerepo='*' && \
-    valkey-server --version | grep -qe "^Server v=$VALKEY_VERSION\." && echo "Found VERSION $VALKEY_VERSION" && \
-    mkdir -p $VALKEY_LIB/data && chown -R valkey:0 $VALKEY_LIB && \
-    mkdir -p $VALKEY_RUN && chown -R valkey:0 $VALKEY_RUN && \
-    chmod -R ug+rwX $VALKEY_RUN && \
-    chmod -R ug+rwX $VALKEY_LIB && \
-    [[ "$(id valkey)" == "uid=1001(valkey)"* ]]
+    redis-server --version | grep -qe "^Redis server v=$REDIS_VERSION\." && echo "Found VERSION $REDIS_VERSION" && \
+    mkdir -p /var/lib/redis/data && chown -R redis.0 /var/lib/redis && \
+    [[ "$(id redis)" == "uid=1001(redis)"* ]]

+# Get prefix path and path to scripts rather than hard-code them in scripts
+ENV CONTAINER_SCRIPTS_PATH=/usr/share/container-scripts/redis \
+    REDIS_PREFIX=/usr \
+    REDIS_CONF=/etc/redis/redis.conf

 COPY root /

@@ -73,7 +65,7 @@
 # script.
 RUN /usr/libexec/container-setup

-VOLUME ["/var/lib/valkey/data"]
+VOLUME ["/var/lib/redis/data"]

 # Using a numeric value because of a comment in [1]:
 # If your S2I image does not include a USER declaration with a numeric user,
@@ -82,4 +74,4 @@
 USER 1001

 ENTRYPOINT ["container-entrypoint"]
-CMD ["run-valkey"]
+CMD ["run-redis"]

Difference between Dockerfile.fedora and Dockerfile.c10s in valkey-container

$ diff -u Dockerfile.c10s Dockerfile.fedora
--- Dockerfile.c10s	2024-08-16 09:53:12
+++ Dockerfile.fedora	2024-08-16 09:53:12
@@ -1,4 +1,4 @@
-FROM quay.io/sclorg/s2i-core-c10s:c10s
+FROM quay.io/fedora/s2i-core:40

 # Valkey image based on Software Collections packages
 #
@@ -25,12 +25,11 @@
       io.k8s.description="$DESCRIPTION" \
       io.k8s.display-name="Valkey $VALKEY_VERSION" \
       io.openshift.expose-services="6379:valkey" \
-      io.openshift.tags="database,valkey,valkey,valkey-$VALKEY_VERSION" \
-      com.redhat.component="valkey-$VALKEY_VERSION-container" \
-      name="sclorg/valkey-$VALKEY_VERSION-c10s" \
+      io.openshift.tags="database,valkey,valkey$VALKEY_VERSION,valkey-$VALKEY_VERSION" \
+      com.redhat.component="$NAME" \
+      name="fedora/$NAME-$VALKEY_VERSION" \
       version="$VALKEY_VERSION" \
-      com.redhat.license_terms="https://www.redhat.com/en/about/red-hat-end-user-license-agreements#rhel" \
-      usage="podman run -d --name valkey_database -p 6379:6379 quay.io/sclorg/valkey-$VALKEY_VERSION-c10s" \
+      usage="podman run -d --name valkey_database -p 6379:6379 quay.io/fedora/$NAME-$VALKEY_VERSION" \
       maintainer="SoftwareCollections.org <[email protected]>"

 EXPOSE 6379
@@ -43,20 +42,20 @@
     VALKEY_LIB=/var/lib/valkey \
     VALKEY_RUN=/run/valkey

-
-# Create user for Valkey that has known UID
+# Create user for valkey that has known UID
 # We need to do this before installing the RPMs which would create user with random UID
 # The UID is the one used by the default user from the parent layer (1001),
 # and since the user exists already, do not create a new one, but only rename
 # the existing
-# This image must forever use UID 1001 for Valkey user so our volumes are
+# This image must forever use UID 1001 for valkey user so our volumes are
 # safe in the future. This should *never* change, the last test is there
 # to make sure of that.
-RUN getent group valkey &> /dev/null || groupadd -r valkey &> /dev/null && \
-    usermod -l valkey -aG valkey -c 'Valkey Server' default &> /dev/null && \
+RUN getent group  valkey &> /dev/null || groupadd -r valkey &> /dev/null && \
+    usermod -l valkey -aG valkey -c 'Server' default &> /dev/null && \
 # Install gettext for envsubst command
-    INSTALL_PKGS="policycoreutils gettext bind valkey" && \
-    dnf install -y --setopt=tsflags=nodocs $INSTALL_PKGS && \
+    dnf install -y dnf-utils gettext policycoreutils && \
+    INSTALL_PKGS="valkey" && \
+    dnf install -y --setopt=tsflags=nodocs --nogpgcheck $INSTALL_PKGS && \
     rpm -V $INSTALL_PKGS && \
     dnf -y clean all --enablerepo='*' && \
     valkey-server --version | grep -qe "^Server v=$VALKEY_VERSION\." && echo "Found VERSION $VALKEY_VERSION" && \
@@ -66,10 +65,6 @@
     chmod -R ug+rwX $VALKEY_LIB && \
     [[ "$(id valkey)" == "uid=1001(valkey)"* ]]

-# Get prefix path and path to scripts rather than hard-code them in scripts
-ENV CONTAINER_SCRIPTS_PATH=/usr/share/container-scripts/valkey \
-    VALKEY_PREFIX=/usr \
-    VALKEY_CONF=/etc/valkey/valkey.conf

 COPY root /

[pkubatrh]

Did a quick sweep and looks fine. But it would be nice to have someone with actual redis/valkey expertise to take a look as well.

[phracek]

[test]

[hhorak]

Well done, @phracek, I didn't see a big issues (just a few left-overs of redis stuff, like v6 version in the imagestream) and think it can serve as a reasonable base like this. A similar conversation about imagestreams might be about templates -- whether there is a reason to maintain templates still, or whether helm charts can replace them.

[phracek]

[test]

[phracek]

@hhorak Helm charts will be added to repository https://github.com/sclorg/helm-charts. Good reminder.

The issue is already filed sclorg/helm-charts#88

[hhorak]

@phracek the imagestreams still don't look correct -- I'd say we should either remove them entirely or convert to valkey, but keeping redis mentions there does not add any value IMO. Otherwise LGTM.

[phracek]

@phracek the imagestreams still don't look correct -- I'd say we should either remove them entirely or convert to valkey, but keeping redis mentions there does not add any value IMO. Otherwise LGTM.

I totally forget to remove it. It would be nice so that tool, who generates sources removes also json's that should not be preset. I will update it.

[phracek]

[test]