In today's episode of kidding on SystemD, we have a surprise guest star appearance - Apache Foundation HTTPD server, or as we in the Debian ecosystem call it, the Apache webserver!

So, imagine a situation like this - Its friday afternoon, you have just migrated a bunch of web domains under a new, up to date, system. Everything works just fine, until... You try to generate SSL certificates from Lets Encrypt.

Such a mundane task, done more than a thousand times already... Yet... No matter what you do, nothing works. Apache just returns a HTTP status code 403 - Forbidden.

Of course, what many folk would think of first when it came to a 403 error is - Ooooh, a permission issue somewhere in the directory structure!

So you check it... And re-check it to make sure... And even switch over to the user the webserver runs under, yet... You can access the challenge just fine, what the hell!

So you go deeper... And enable the most verbose level of logging apache is capable of - Trace8. That tells you... Not a whole lot more... Apparently, the webserver was unable to find file specified? But... Its right there, you can see it!

So you go another step deeper and start tracing the process' system calls to see exactly where it calls stat/lstat on the file, and you see that it... Calls lstat and... It... Returns -1? What the hell#2!

So, you compile a custom binary that calls lstat on the first argument given and prints out everything it returns... And... It works fine!

Until now, I chose to omit one important detail that might have given away the issue to the more knowledgeable right away. Our webservers have the URL /.well-known/acme-challenge/, used for ACME challenges, aliased somewhere else on the filesystem - To /tmp/challenges.

See the issue already?

Some *bleep* over at the Debian Package Maintainer group decided that Apache could save very sensitive data into /tmp, so, it would be for the best if they changed something that worked for decades, and enabled a SystemD service unit option "PrivateTmp" for the webserver, by default.

What it does is that, anytime a process started with this option enabled writes to /tmp/*, the call gets hijacked or something, and actually makes the write to a private /tmp/something/tmp/ directory, where something... Appeared as a completely random name, with the "apache2.service" glued at the end.

That was also the only reason why I managed fix this issue - On the umpteenth time of checking the directory structure, I noticed a "systemd-private-foobarbas-apache2.service-cookie42" directory there... That contained nothing but a "tmp" directory with 777 as its permission, owned by the process' user and group.

Overriding that unit file option finally fixed the issue completely.

I have just one question - Why? Why change something that worked for decades? I understand that, in case you save something into /tmp, it may be read by 3rd parties or programs, but I am of the opinion that, if you did that, its only and only your fault if you wrote sensitive data into the temporary directory.

And as far as I am aware, by default, Apache does not actually write anything even remotely sensitive into /tmp, so...

Why. WHY!
I wasted 4 hours of my life debugging this! Only to find out its just another SystemD-enabled "feature" now!

And as much as I love kidding on SystemD, this time, I see it more as a fault of the package maintainers, because... I found no default apache2/httpd service file in the apache repo mirror... So...

In last episode of "How SystemD screwed me over", we talked about Systemd's PrivateTMP and how it stopped me from generating SSL certificates.

In today's episode - SystemD vs CGroups!

Mister Pottering and his team apparently felt that CGroups are underused (As they can be quite difficult to set up), and so decided to integrate them into SystemD by default. As well as to provide a friendlier interface to control their values.

One can read about these interactions in the manual page "systemd.resource-control"

All is cool so far. So what happened to me today?

Imagine you did a major system release upgrade of a production server, previously tested on a standalone server. This upgrade doesn't only upgrade the distribution however, it also includes the switch from SysVInit to SystemD. Still, everything went smooth before, nothing to worry now then, right? Wrong.

The test server was never properly stress-tested. This would prove to be an issue.

When the upgrade finishes, it is 4 AM. I am happy to go to bed at last. At 6 AM, however, I am woken up again as the server's webservices are unavailable, and the machine is under 100% CPU load. Weird, I check htop and see that Apache now eats up all 32 virtual cores. So I restart it, casting it off to some weird bug or something as the load returns to normal.

2 hours later, however, the same situation occurs. This time, I scour all the logs I can, and find something weird - Many mentions that Apache couldn't create a worker thread? That's weird.

Several hours of research and tinkering later, I found out the following:
1 - By default, all processes of a system that runs SystemD are part of several CGroups. One of these CGroups is the PID CGroup, meant to stop a runaway process from exhausting all PIDs/TIDs of a system.

This limit is, by default, set to a certain amount of the total available PIDs. If a process exhausts this limit, it can no longer perform operations like fork().

So now, I know the how and why, but how should I solve this? The sanest option would be to get a rough estimate of just how many threads the Apache webserver might need. This option, though, is harder, than apparent. I cannot just take the MaxRequestsWorkers number... The instance has roughly double the amount of threads already. The cause being, as I found out, the HTTP/2 module, which spawns additional threads that do not count towards this limit. So I have no idea what limit to set.

Or I could... Disable the limit for just the webserver via the TasksAccounting switch. I thought this would work. And it did seem to... Until I ran out of TIDs again - Although systemctl status apache2.service no longer reported the number of tasks or a task limit of the process, the PID CGroup stayed set to the previous limit. Later I found out that I can only really disable the Task Accounting for all the units of a given slice and its parents.

This, though, systemctl somewhat didn't make apparent (And I skimmed the manual, that part was my fault)

So... The only remaining option I had was to... Just set the limit to infinite. And that worked, at last.

It took me several hours to debug this issue. And I once again feel like uninstalling systemd again, in favor of sysvinit.

What did I learn? RTFM, carefully, everything is important, it is not enough to read *half* the paragraph of a given configuration option...

Oh, and apache + http/2 = huge TID sink.