Author: Roo

Knowing when to update your docker containers with DIUN

DIUN – Docker Image Update Notifier. I was very glad to come across this particular tool as it helped solve a problem I had, one that I felt strongly enough about that I’d put a bunch of time into creating something similar.

My approach, was to build some scripting to determine the signature of the image that I had deployed locally, and then make many queries to the registry to determine what (if any) changes were in the remote image. This immediately ran into some of the API limits on dockerhub. There were also other challenges with doing what I wanted. The digest information you get with docker pull doesn’t match the digest information available on dockerhub. I did fine this useful blog post (and script) that solves a similar problem, but also hits some of the same API limitations. It seemed like maybe a combination of web scraping plus API calls could get a working solution, but it was starting to be a hard problem.

DIUN uses a very different approach. It starts by figuring out what images you want to scan – the simplest way to do this is to allow it to look at all running docker containers on your system. With this list of images, it can then query the docker image repository for the tag of that image. On the first run, it just saves this value away in a local data store. Every future run, it compares the tag it fetched to the one in the local data store – if there is a difference, it notifies you.

In practice, this works to let you know every time a new image is available. It doesn’t know if you’ve updated your local image or not, nor does it tell you what changed in the image – only that there is a newer version. Still, this turns out to be quite useful especially when combined with slack notifications.

Setting up DIUN for my system was very easy. Here is the completed Makefile based on my managing docker container with make post.

I started very simply at first following the installation documentation provided. I used a mostly environment variable approach to configuring things as well. The three variables I need to get started were:

  • DIUN_WATCH_SCHEDULE – enable cron like behaviour
  • DIUN_PROVIDERS_DOCKER – watch all running docker containers
  • DIUN_PROVIDERS_DOCKER_WATCHBYDEFAULT – watch all by default

Looking at the start-up logs for the diun container is quite informative and generally useful error messages are emitted if you have a configuration problem.

I later added the:

  • DIUN_NOTIF_SLACK_WEBHOOKURL

in order to get slack based notifications. There is a little bit of setup you need to do with your slack workspace to enable slack webhooks to work, but it is quite handy for me to have a notification in a private channel to let me know that I should go pull down a new container.

Finally I added a configuration file ./data/config.yml to capture additional docker images which are used as base images for some locally built Dockerfiles. This will alert me when the base image I’m using gets an update and will remind me to go re-build any containers that depend on them. This use the environment varible:

  • DIUN_PROVIDERS_FILE_FILENAME

My configuration file looks like:

I’ve actually been running with this for a couple of weeks now. I really like the linuxserver.io project and recommend images built by them. They have a regular build schedule, so you’ll see (generally) weekly updates for those images. I have nearly 30 different containers running, and it’s interesting to see which ones are updated regularly and which seem to be more static (dormant).

Some people make use of Watchtower to manage their container updates. I tend to subscribe to the philosophy that this is not a great idea for a ‘production’ system, at least some subset of the linuxserver.io folks agree with this as well. I like to have hands on keyboard when I do an update, so I can make sure that I’m around to deal with any problems that may happen.

 

Hacking an old HP Chromebook 11 G5

When it was time to get one of the kids a Chromebook for school years ago, I made sure to purchase a 4GB memory model with an Intel chip. I’m a fan of ARM devices, but at the time (6+ years ago) there was some real junk out there. There was also a price factor and I was looking at the lower end market, durability was also a concern.

I remember the Dell Chromebook 11″ was a hot item back then, but the pricing was higher than I wanted. Same for the Lenovo Chromebooks. After a bunch of searching around I found a nice HP Chromebook 11 G5 (specs) – if my memory is correct I got this well under $300 at the time.

This HP 11 G5 worked well, survived a few drops, and made it until it’s end of life – when Google stops providing OS updates. I’ve since replaced it with a nice Lenovo IdeaPad Flex 5 Chromebook – a nice step up, and there was a refurb model available for a great price (under $350).

For a long time there has been Neverware CloudReady – a neat way to get ChromeOS on old laptops. I always worried that there were security concerns with some random company offering ‘Google’ logins, but Neverware worked well. Google has since bought CloudReady, and seems to have turned around and created Chromeos Flex as the successor.

I figured that I could use Chromeos Flex on the HP 11 G5 to continue to get updates. Another solution would be to look at turning it into a GalliumOS machine. I actually have another old 14″ Chromebook I have run GalliumOS on, but have since moved to Linux Mint and use it as a generic Linux laptop.

I would recommend reading through the GalliumOS wiki information carefully to learn about the process of converting a Chromebook into a useful generic low end laptop. Specifically the Preparing section, a review of the Hardware Compatibility section and Firmware sections. Inevitably you’ll also end up on MrChromeBox’s site – which is where you’ll get the firmware replacement you’ll need.

While you can in some cases get alternative firmware running on the Chromebook hardware, it’s much easier if you go remove the hardware write protect. There wasn’t a specific guide to doing this, but the iFixit site was useful for the tear down aspect.

You will want to remove the black screw pointed at by the arrow. It’s near the keyboard ribbon cable connector. This is the hardware write protect.

Once I’d done this, it was simply a matter of installing the
“UEFI (Full ROM) firmware” using the MrChromeBox scripts. This is not for the faint of heart, and I do recommend making a backup of the original firmware in case you want to go back.

At this point you can install any old OS distribution you want. In my case I wanted to install Chromeos Flex, so I’d downloaded that and created a USB drive with it ready to roll. Installing it on my newly firmware updated Chromebook was easy.

I then ran into trouble. While Chromeos starts up fine, it was quickly clear that sound didn’t work. The video camera was working fine, but I couldn’t get any output or input for sound. I found that others had this same issue. I even tried using wired headphones (same problem) and bluetooth headphones (sound out was fine, sound in didn’t work at all)

This is a bummer, but understandable. Chromebook hardware is not really the target for Chromeos Flex. I figured it was worth trying out a generic Linux distro, so I picked Linux Mint. Booting from a USB drive with Mint on it was again easy with the new firmware. Sound output worked fine, as did web cam video – but the mic was still a problem, again something others had discovered.

At this point Chromeos Flex was a dead end. I can’t give someone a Chromebook that doesn’t have audio in or out and no reasonable work-arounds to get there. Installing Linux won’t trivially solve the problem because I get sound out, but no mic.

Remember when I said it was a good idea to backup the original firmware? Yup, we’re returning this Chromebook to stock (but I’ll leave the write protect screw out – because why not?). The MrChromeBox FAQ walks you through restoring that firmware. Since I had Linux Mint on a bootable USB I just used that to start up a shell and pull the script. Once I’d restored the stock firmware, I needed to build a ChromeOS recovery image and then return to a totally stock setup.

Now this old HP 11 G5 Chromebook has all of it’s features working, video, sound, mic.. but is trapped on an expired version of ChromeOS. Eventually the browser will become annoyingly old and at that point you’ll have to decide between the limitations of the browser, or losing your mic (and possibly sound).

When rate limiting (and firewalling) goes wrong


Recently I experienced a few power failures that lasted hours. This means that when the power is back, all of my infrastructure reboots and reconnects. For the most part this is 100% automatic, but the last time I ran into an interesting problem.

My pi-hole was running with the default rate limiting of 1000/60. This means that each device can make up to 1000 requests per minute, and if it exceeds that it will be put on a deny list for 60 seconds.

It turns out that my main server that runs a bunch of docker containers makes a lot of DNS requests when everything is starting up all at once. This creates a storm of requests to the pi-hole and the server ends up being blocked for DNS requests (responding with REFUSED) due to rate limiting.

Unfortunately the behaviour of enough of the containers is to retry when this happens. This causes more DNS requests to be made as the retry logic runs. These retries cause another wave of requests which cause the server to be blocked again. Some of my containers entered error conditions due to unexpected DNS failures, so these needed to later be restarted but at least they stopped contributing to the problem.

My email container was pretty unhappy, it really wants to be able to use DNS, even when receiving email. Since my server had been unavailable for a while, there were external email servers trying to deliver mail that had been queued – this contributed to the load. Additionally I couldn’t connect any email clients to the server which left me scratching my head a little, more on that later on.

The ‘fix’ was easy enough. Modify the pi-hole DNS rate-limiting setting to 0/0 to remove any rate limiting. This is imperfect, but at one point I saw 30,000 requests in a minute from my struggling server and I think I’d rather have no limit and deal with that problem than hit the limit and run into this denial of service issue.

Now that the pi-hole was happy, I was able to get most of my containers to be happy with a little poking at them. Email was still sad, and this took me a coffee break to realize what was wrong. The email container was receiving email just fine, but I could not connect with a client. This felt like a networking problem, but how could that be?

I had forgotten (again) – that the email server has fail2ban running in it. This scans logs looking for suspicious activity and will ban an IP for a period of time by inserting a firewall rule. Furthermore, as I use the domain name to configure my email client – this resolves to the external IP. The external IP means that the client talks to my OpenWRT router which provides NAT and then redirects/maps that external IP back into my network. This has the effect that the originating IP looks like it is my router, not the client machine on the internal IP address. This process is called NAT reflection, or NAT hairpinning.

While NAT reflection is a super handy feature for my OpenWRT router to have, allowing me to easily from inside my home network visit a machine I’ve exposed via port mapping to the outside world using the same DNS entry that points at the external IP address — it means that services on that machine see my router IP as the client IP. When any of the machines in my house have problems connecting to my email server, in this case because I had DNS REFUSED errors on the email server, fail2ban decides that is a bad client and bans it. Thus banning all traffic originating from my home network.

This is easy to fix once you understand what is happening, I just needed to unban my router IP and my email clients could connect.