Kangie

joined 1 year ago
[–] [email protected] 4 points 11 months ago

You could probably fit something inside the chassis, but why? It doesn't have standard mounting hardware for any ATX form factor, you'll need to source a slim PSU, and at the end of the day, without a ton of effort to make the build work, and additional thought put into the cooling (etc) you're going to destroy a working console for a pretty average result with a lot of trade-offs.

TL;DR, bad idea. You can do better.

[–] [email protected] 5 points 11 months ago

Don't worry too much about it: it's not the 80s and we're not the Soviets.

Once it's in orbit it doesn't matter and if there's issues during ascent the source will be hardened to prevent catastrophic release.

Realistically though, just send it up on a falcon 9, the track record on those things is _ astounding_.

[–] [email protected] 13 points 11 months ago

It depends. I'm glad that we have tools like proton, but when this was an explicit stretch goal that was met during funding it's a bit different.

I'm disappointed in Nightdive. Not just because they cancelled the port, but because they made a promise and they broke it, and they just remained silent about it instead of being transparent about any challenges that they were facing over the preceding years.

[–] [email protected] 3 points 11 months ago

The same day as they announce they're not doing the system shock native client eh?

[–] [email protected] 6 points 1 year ago (1 children)

I may be a touch biased, but I feel like you might enjoy trying Gentoo one day, especially with the recent official binary package host.

[–] [email protected] 2 points 1 year ago (2 children)

Could you elaborate?

[–] [email protected] 18 points 1 year ago* (last edited 1 year ago) (2 children)

TL;DR don't worry (for now) - it only impacts rpm and deb builds and impacted releases only really made it into OpenSuSe tumbleweed - if you're running bleeding edge maybe you need to worry a little.

A laymans explanation about what happens is that the malicious package uses an indirect linkage (via systemd) to openssh and overrides a crypto function which either:

  • allows access to the system to a particular key
  • allows remote code execution with a particular key

Or both!

I have secondhand info that privately the reverse engineering is more advanced, but nobody wants to lead with bad info.

As for what you should do? Unless you're running an rpm or deb based distro and you have version 5.6.0 or 5.6.1 of xz-utils installed, not much. If you are, well, that comes down to your threat model and paranoia level: either upgrade (downgrade) the package to a non-vulnerable version or dust off and nuke the site from orbit; it's the only way to be sure.

[–] [email protected] 1 points 1 year ago

Wow. What fucking garbage.

[–] [email protected] 27 points 1 year ago

there is no good answer

There is clearly a worst option.

[–] [email protected] 7 points 1 year ago (3 children)

I use Traefik for all of my containerised services. It's fantastic.

[–] [email protected] 50 points 1 year ago (3 children)

I'm sorry, but no. PluralKit only really impacts a tiny minority of the userbase to begin with. It isn't enough to cause people outside that group to choose the platform, nor is it enough for people outside of that minority to avoid moving to whatever the next big thing is.

 

cross-posted from: https://lemmy.srcfiles.zip/post/2231272

I've been interested in switching away from $PROPRIETARY_VENDOR's HPC node / cluster management offering for a while, and the opportunity has finally arisen - $VENDOR has decided to massively hike up their prices, so it's time to look at alternatives.

The top option on my list is Warewulf. Warewulf is a stateless node management tool, where stateless means "we boot any image you want into memory" (compared with competing implementations which do 'magic' to image a node's disk every boot). There are advantages and disadvantages to each approach.

The thing that attracts me most to Warewulf is that they've come to the conclusion that most HPC "disk images" are basically container images. Rather than using a chroot directory as an image (as do so many competing implementations) Warewulf have leaned wholeheartedly into the concept, and have adopted the OCI image tooling to define HPC images!

This offers an astounding amount of flexibility that the current $VENDOR solution does not - the ability to define, build, and run any (reasonable) flavour of Linux as an image for HPC nodes; images need only the kernel, networking, systemd, and (optionally) a nfs client (this is for convenience, it's not required for node functionality).

Based on that I've taken it upon myself to have some fun and investigate the current state of Warewulf as a node management tool - the first step, of course, was creating a Gentoo ebuild for Warewulf that compiles and installs. I'm happy to say that, after fixing some bugs in the offline build process, I have a working ebuild.

If you're a Gentooer with an interest in Node Management or HPC, please give my Warewulf ebuild a try; it compiles and installs but I haven't yet had a chance to do any real cluster management and I'm interested in hearing about any bugs you encounter!

Next Week(-ish): Gentoo HPC Base Images - I'm going to have to dig into Gentoo Networking to update the wwinit image.

 

cross-posted from: https://lemmy.srcfiles.zip/post/32334

What is Kubler?

Kubler is a generic, extendable build orchestrator, written in Bash. It can be used to take advantage of Portage’s features to build lightweight Docker or Podman images without needing to mess with crossdev, or as a tool to assist with ebuild development.

Why should you use it?

  • You like lightweight, easy-to-create, containers
  • You want to reduce the attack surface by including only what’s required
  • You want to take advantage of USE flags to manage package features
  • You want the awesome package library offered by the Gentoo ebuild repository (and other ebuild repos)
  • You want up-to-date containers
  • You want a containerised environment for building and testing ebuilds

A real-world example

I recently needed to integrate a containerised application with a vendor-managed openldap instance that uses mTLS authentication. Unfortunately the containerised application does not work with mTLS and the vendor managed openldap instance can’t be easily configured to use anything else.

I came up with the solution of using openldap’s lloadd LDAP load balancer daemon to proxy connections from an encrypted internal network to the LDAP server but was left with the issue that I didn’t have a working openldap container that contained lloadd - of the existing containers that I tried the only one that actually had an lloadd bin didn’t actually include required dependencies!

Glossing over a recent ebuild update to openldap to enable the building of lloadd, enter Kubler - It’s turned out to be an incredibly flexible and hands-off tool compared to trying to accomplish the same thing with (e.g.) Dockerfiles.

Kubler in action

This (lightly sanitised) real-world example creates create a new namespace called ‘larry’ which may contain multiple images.

Use the new command to take care of the boilerplate; choose ‘multi’ when asked for the namespace type:`

$ kubler new namespace larry
»»»
»»»  to accept default value
»»»
»»» Working dir type? Choices:
»»»   single - You can't add further namespaces to the created working dir, it only holds images
»»»   multi  - Creates a working dir that can hold multiple namespaces
»[?]» Type (single): multi
»»»
»»» Top level directory name for new namespace 'larry'? The directory is created at /data/development/gentoo-containers/
»[?]» Namespaces Dir (kubler-images):
»»»
»»»»» Initial image tag, a.k.a. version?
»[?]» Image Tag (20230706):
»»»
»[!]» New namespace location:  /data/development/gentoo-containers/kubler-images/larry
»»»
»»»»» Who maintains the new namespace?
»[?]» Name (Your Name): Larry the Cow
»[?]» EMail ([email protected]): [email protected]
»»»
»»»»» Default build engine?
»[?]» Engine (docker):
»»»
»[✔]» Successfully created "larry" namespace at /data/development/gentoo-containers/kubler-images
»»»
»[!]» Configuration file: /data/development/gentoo-containers/kubler-images/larry/kubler.conf
»»»
»[!]» To manage the new namespace with GIT you may want to run:
»»»
»»» $ git init /data/development/gentoo-containers/kubler-images/larry
»»»
»[!]» To create images in the new namespace run:
»»»
»»» $ cd /data/development/gentoo-containers/kubler-images/larry
    $ kubler new image larry/

Although not strictly required, installing Kubler’s example images is a good idea.

$ cd larry/
$ kubler update

It is worthwhile to begin tracking this new namespace with Git so that images can be tracked as they are created and updated. Kubler has already placed a prepopulated a .gitignore file for convenience.

pushd /data/development/gentoo-containers/kubler-images/larry
git init .
git add .
git commit -m "Initial commit"
popd

Create the new ‘openldap’ within the existing ‘larry’ namespace, based on the ‘kubler/busybox’ image.

kubler new image larry/openldap
»»»
»»»  to accept default value
»»»
»»» Extend an existing Kubler managed image? Fully qualified image id (i.e. kubler/busybox) or scratch
»[?]» Parent Image (scratch): kubler/busybox
»»»
»»» Add test template(s)? Possible choices:
»»»   hc  - Add a stub for Docker's HEALTH-CHECK, recommended for images that run daemons
»»»   bt  - Add a stub for a custom build-test.sh script, a good choice if HEALTH-CHECK is not suitable
»»»   yes - Add stubs for both test types
»»»   no  - Fck it, we'll do it live!
»[?]» Tests (hc): yes
»»»
»[✔]» Successfully created new image at /data/development/gentoo-containers/kubler-images/larry/images/openldap
»»»

Note: This step is ‘‘not’’ required; it is possible to directly edit the build.sh file if you are familiar with Portage.

Kubler brings a unique feature to the table when constructing an container image: the --interactive build argument. As the name implies, this launches the build container in an interactive manner, enabling users to investigate the current / inherited configuration.

$ kubler build larry/openldap -i

This will build any missing parent images/builders; the first run may take quite a bit of time - once the local binary package cache and build containers are seeded future runs will be much faster. Once the prerequisite images are ready the build container will present a shell.

For first-time users it may be convenient to search for the openldap package to ensure that the correct atom is selected and investigate any USE flags that are of interest:

# eix openldap
* net-nds/openldap
     Available versions:  2.4.59-r2^t 2.5.14(0/2.5)^t 2.6.3-r7(0/2.6)^t ~2.6.4-r1(0/2.6)^t ~2.6.4-r2(0/2.6)^t {argon2 autoca +berkdb +cleartext crypt cxx debug experimental gnutls iodbc ipv6 kerberos kinit minimal odbc overlays pbkdf2 perl samba sasl selinux sha2 smbkrb5passwd ssl static-libs +syslog systemd tcpd test ABI_MIPS="n32 n64 o32" ABI_S390="32 64" ABI_X86="32 64 x32"}
     Homepage:            https://www.openldap.org/
     Description:         LDAP suite of application and development tools

Edit the image’s build script:

nano /config/build.sh

Note: The /config directory in the build container is the host mounted image directory at larry/images/openldap/. Feel free to use a local IDE/editor to edit build.sh instead.

Add the net-nds/openldap and net-misc/curl packages to the _packages variable in build.sh, update cURL USE flags, enable the ~arch (~amd64 - the Gentoo ‘testing’ keyword) for packages we care about:

_packages="net-nds/openldap net-misc/curl"
...
configure_rootfs_build()
{
    # Update a Gentoo package use flag.
    update_use 'net-misc/curl' '+ldap'
    # ..or a Gentoo package keyword
    update_keywords 'net-misc/curl' '+~amd64'
    update_keywords 'net-nds/openldap' '+~amd64'
...
}

Note: If using the busybox image as a parent, unset the su USE flag from sys-apps/util-linux in the build.sh.

Perform a test run of the first build phase:

$ kubler-build-root

Once this completes successfully exit the interactive builder using exit.

Building the image

Assuming that build.sh has been configured as described above, it should be safe to attempt to build the image.

$ kubler build larry/openldap -nF
»[✘]»[larry/openldap]» fatal: build-test.sh for image larry/openldap:20230704 failed with exit signal: 1

Note: The arguments are short hand for --no-deps and --force-full-image-build, omitting -n would also rebuild all parent images, which is waste of time in this case.

The build will fail, as expected, due to the build-test.sh script not being implemented. This is a good time to implement the build-test.sh script, which will be used to verify that the image is functional.

Note: pipefail will cause build-test.sh to fail on busybox-based images

#!/usr/bin/env sh

set -eo

# Do some tests and exit with either 0 for healthy or 1 for unhealthy
# Check that the openldap bin launches and provides some expected output
/usr/lib/openldap/lloadd -VV  2>&1 | grep "OpenLDAP" || exit 1

exit 0

Unfortunately this image is not suitable for a build-time docker health check via the docker-healthcheck.sh mechanism, so must be disabled in larry/images/openldap/build.conf:

POST_BUILD_HC=false

A health check suitable for your environment should be provided using standard docker syntax in the image’s Dockerfile.template instead. Ensure that the provided docker-healthcheck.sh script iS updated (or commented out of the dockerfile) as the default will fail.

Modify the image’s Dockerfile.template to add any finishing touches, such as the ENTRYPOINT or CMD directives. In this example the container will act as an LDAP proxy via lloadd; additional configuration will be provided at runtime by mounting the configuration into the container.

FROM ${IMAGE_PARENT}
LABEL maintainer="${MAINTAINER}"

ADD rootfs.tar /

COPY docker-healthcheck.sh /usr/bin/docker-healthcheck
HEALTHCHECK --interval=60s --timeout=5s --start-period=5s --retries=3 CMD ["docker-healthcheck"]

CMD ["/usr/lib/openldap/lloadd"]

Re-run the build:

$ kubler build larry/openldap -nF
»[✔]»[larry/openldap]» done.

At this point the image should exist in the local Docker/Podman registry and be ready for use:

docker images
REPOSITORY                                              TAG                       IMAGE ID       CREATED          SIZE
larry/openldap                                          20230704                  09347c55282b   2 minutes ago    56.4MB
larry/openldap                                          latest                    09347c55282b   2 minutes ago    56.4MB

Hopefully this has been useful and you are now ready to build your own images! I’ve been incredibly impressed with how easy it is to use the tool (and it it’ll run from any distro with a recent version of Docker/Podman), and the quality of the resulting images. I’m a recent convert, but have updated the Gentoo Wiki with the above information (and some extra info on using it for ebuild development) and will be using Kubler in future wherever I need to create images.

Happy containering!

 

cross-posted from: https://lemmy.srcfiles.zip/post/3841

What is Kubler?

Kubler is a generic, extendable build orchestrator, written in Bash. It can be used to take advantage of Portage's features to build lightweight Docker or Podman images without needing to mess with crossdev, or as a tool to assist with ebuild development.

Why should you use it?

  • You like lightweight, easy-to-create, containers
  • You want to reduce the attack surface by including only what's required
  • You want to take advantage of USE flags to manage package features
  • You want the awesome package library offered by the Gentoo ebuild repository (and other ebuild repos)
  • You want up-to-date containers
  • You want a containerised environment for building and testing ebuilds

A real-world example

I recently needed to integrate a containerised application with a vendor-managed openldap instance that uses mTLS authentication. Unfortunately the containerised application does not work with mTLS and the vendor managed openldap instance can't be easily configured to use anything else.

I came up with the solution of using openldap's lloadd LDAP load balancer daemon to proxy connections from an encrypted internal network to the LDAP server but was left with the issue that I didn't have a working openldap container that contained lloadd - of the existing containers that I tried the only one that actually had an lloadd bin didn't actually include required dependencies!

Glossing over a recent ebuild update to openldap to enable the building of lloadd, enter Kubler - It's turned out to be an incredibly flexible and hands-off tool compared to trying to accomplish the same thing with (e.g.) Dockerfiles.

Kubler in action

This (lightly sanitised) real-world example creates create a new namespace called 'larry' which may contain multiple images.

Use the new command to take care of the boilerplate; choose 'multi' when asked for the namespace type:`

$ kubler new namespace larry
»»»
»»» <enter> to accept default value
»»»
»»» Working dir type? Choices:
»»»   single - You can't add further namespaces to the created working dir, it only holds images
»»»   multi  - Creates a working dir that can hold multiple namespaces
»[?]» Type (single): multi
»»»
»»» Top level directory name for new namespace 'larry'? The directory is created at /data/development/gentoo-containers/
»[?]» Namespaces Dir (kubler-images):
»»»
»»»»» Initial image tag, a.k.a. version?
»[?]» Image Tag (20230706):
»»»
»[!]» New namespace location:  /data/development/gentoo-containers/kubler-images/larry
»»»
»»»»» Who maintains the new namespace?
»[?]» Name (Your Name): Larry the Cow
»[?]» EMail ([email protected]): [email protected]
»»»
»»»»» Default build engine?
»[?]» Engine (docker):
»»»
»[✔]» Successfully created "larry" namespace at /data/development/gentoo-containers/kubler-images
»»»
»[!]» Configuration file: /data/development/gentoo-containers/kubler-images/larry/kubler.conf
»»»
»[!]» To manage the new namespace with GIT you may want to run:
»»»
»»» $ git init /data/development/gentoo-containers/kubler-images/larry
»»»
»[!]» To create images in the new namespace run:
»»»
»»» $ cd /data/development/gentoo-containers/kubler-images/larry
    $ kubler new image larry/<image_name>

Although not strictly required, installing Kubler's example images is a good idea.

$ cd larry/
$ kubler update

It is worthwhile to begin tracking this new namespace with Git so that images can be tracked as they are created and updated. Kubler has already placed a prepopulated a .gitignore file for convenience.

pushd /data/development/gentoo-containers/kubler-images/larry
git init .
git add .
git commit -m "Initial commit"
popd

Create the new 'openldap' within the existing 'larry' namespace, based on the 'kubler/busybox' image.

kubler new image larry/openldap
»»»
»»» <enter> to accept default value
»»»
»»» Extend an existing Kubler managed image? Fully qualified image id (i.e. kubler/busybox) or scratch
»[?]» Parent Image (scratch): kubler/busybox
»»»
»»» Add test template(s)? Possible choices:
»»»   hc  - Add a stub for Docker's HEALTH-CHECK, recommended for images that run daemons
»»»   bt  - Add a stub for a custom build-test.sh script, a good choice if HEALTH-CHECK is not suitable
»»»   yes - Add stubs for both test types
»»»   no  - Fck it, we'll do it live!
»[?]» Tests (hc): yes
»»»
»[✔]» Successfully created new image at /data/development/gentoo-containers/kubler-images/larry/images/openldap
»»»

Note: This step is ''not'' required; it is possible to directly edit the build.sh file if you are familiar with Portage.

Kubler brings a unique feature to the table when constructing an container image: the --interactive build argument. As the name implies, this launches the build container in an interactive manner, enabling users to investigate the current / inherited configuration.

$ kubler build larry/openldap -i

This will build any missing parent images/builders; the first run may take quite a bit of time - once the local binary package cache and build containers are seeded future runs will be much faster. Once the prerequisite images are ready the build container will present a shell.

For first-time users it may be convenient to search for the openldap package to ensure that the correct atom is selected and investigate any USE flags that are of interest:

# eix openldap|output=<pre>
* net-nds/openldap
     Available versions:  2.4.59-r2^t 2.5.14(0/2.5)^t 2.6.3-r7(0/2.6)^t ~2.6.4-r1(0/2.6)^t ~2.6.4-r2(0/2.6)^t {argon2 autoca +berkdb +cleartext crypt cxx debug experimental gnutls iodbc ipv6 kerberos kinit minimal odbc overlays pbkdf2 perl samba sasl selinux sha2 smbkrb5passwd ssl static-libs +syslog systemd tcpd test ABI_MIPS="n32 n64 o32" ABI_S390="32 64" ABI_X86="32 64 x32"}
     Homepage:            https://www.openldap.org/
     Description:         LDAP suite of application and development tools

Edit the image's build script:

nano /config/build.sh

Note: The /config directory in the build container is the host mounted image directory at larry/images/openldap/. Feel free to use a local IDE/editor to edit build.sh instead.

Add the net-nds/openldap and net-misc/curl packages to the _packages variable in build.sh, update cURL USE flags, enable the ~arch (~amd64 - the Gentoo 'testing' keyword) for packages we care about:

_packages="net-nds/openldap net-misc/curl"
...
configure_rootfs_build()
{
    # Update a Gentoo package use flag.
    update_use 'net-misc/curl' '+ldap'
    # ..or a Gentoo package keyword
    update_keywords 'net-misc/curl' '+~amd64'
    update_keywords 'net-nds/openldap' '+~amd64'
...
}

Note: If using the busybox image as a parent, unset the su USE flag from sys-apps/util-linux in the build.sh.

Perform a test run of the first build phase:

$ kubler-build-root

Once this completes successfully exit the interactive builder using exit.

Building the image

Assuming that build.sh has been configured as described above, it should be safe to attempt to build the image.

$ kubler build larry/openldap -nF
»[✘]»[larry/openldap]» fatal: build-test.sh for image larry/openldap:20230704 failed with exit signal: 1

Note: The arguments are short hand for --no-deps and --force-full-image-build, omitting -n would also rebuild all parent images, which is waste of time in this case.

The build will fail, as expected, due to the build-test.sh script not being implemented. This is a good time to implement the build-test.sh script, which will be used to verify that the image is functional.

Note: pipefail will cause build-test.sh to fail on busybox-based images

#!/usr/bin/env sh

set -eo

# Do some tests and exit with either 0 for healthy or 1 for unhealthy
# Check that the openldap bin launches and provides some expected output
/usr/lib/openldap/lloadd -VV  2>&1 | grep "OpenLDAP" || exit 1

exit 0

Unfortunately this image is not suitable for a build-time docker health check via the docker-healthcheck.sh mechanism, so must be disabled in larry/images/openldap/build.conf:

POST_BUILD_HC=false

A health check suitable for your environment should be provided using standard docker syntax in the image's Dockerfile.template instead. Ensure that the provided docker-healthcheck.sh script iS updated (or commented out of the dockerfile) as the default will fail.

Modify the image's Dockerfile.template to add any finishing touches, such as the ENTRYPOINT or CMD directives. In this example the container will act as an LDAP proxy via lloadd; additional configuration will be provided at runtime by mounting the configuration into the container.

FROM ${IMAGE_PARENT}
LABEL maintainer="${MAINTAINER}"

ADD rootfs.tar /

COPY docker-healthcheck.sh /usr/bin/docker-healthcheck
HEALTHCHECK --interval=60s --timeout=5s --start-period=5s --retries=3 CMD ["docker-healthcheck"]

CMD ["/usr/lib/openldap/lloadd"]

Re-run the build:

$ kubler build larry/openldap -nF
»[✔]»[larry/openldap]» done.

At this point the image should exist in the local Docker/Podman registry and be ready for use:

docker images
REPOSITORY                                              TAG                       IMAGE ID       CREATED          SIZE
larry/openldap                                          20230704                  09347c55282b   2 minutes ago    56.4MB
larry/openldap                                          latest                    09347c55282b   2 minutes ago    56.4MB

Hopefully this has been useful and you are now ready to build your own images! I've been incredibly impressed with how easy it is to use the tool (and it it'll run from any distro with a recent version of Docker/Podman), and the quality of the resulting images. I'm a recent convert, but have updated the Gentoo Wiki with the above information (and some extra info on using it for ebuild development) and will be using Kubler in future wherever I need to create images.

Happy containering!

 

A neofetch screenshot from the machine I just unpacked and am finally ready to do actual work on. I swear I have a 6.4 kernel pending a reboot!

view more: next ›