Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
-
Two days ago I had a particularly retarded conversation with my father.
Father: Have you heard of [insert some language that will probably die in half a year]?
Me: Never heard of it.
Father: I talked with a friend yesterday and he said people pay millions for it.
Me: And?
Father: You have to learn it.
Me: *At this point trying not to facepalm* I don't think that's a good idea.
Father: Why?
Me: *begins explaining how the industry changes quickly, underused languages die out or how there simply aren't many materials on them*
My father then started shouting at me, saying that I don't listen to him and that his friend (who does stock trading for living) knows what's trending right now. Couple that with the fact that I was sick in bed when this random event happened, it led to me wanting to, ever so slightly, fucking kill myself.5 -
There's a local server with lots of processing power and plenty of GPUs and tons RAM going critically underused. Why? There's someone who's running a process using a relational database which literally performs 40 to 60 percentage disk I/O for months. It's so bad, if you run "ls" it may take a minute to run (only if you are lucky to ssh in).
4 -
the one that exists (c#) seems underused compared to where it could (or even should) be used. and the place that uses it the most (enterprise) butchers and mangles its use, just as enterprise tends to do with everything.
the one that i'm designing... the fact that it doesn't exist yet, and that even as i'm zeroing in on syntax and philosophy that i'm very much starting to be proud of, i still don't have a proper idea of how to implement even the most basic parser/interpreter for it, not because it's in any way difficult or unusual, but just because... i've never done that before, so i get into weird circular thought paths that produce weird nonsensical code...
... on top of that, i still only have a very, very fuzzy idea of how will it (sometime in extremely distant future) actually implement the most interesting and core feature - event-based continuous (partial) re-parsing of the source code and the fact that traversing the tokens at the leaf level of the syntax tree should result in valid machine code (or at least assembly) that is the "compiled" program.
i *know* it's possible, i just don't yet know enough to have a contrete idea how exactly to achieve it.
but imagine - a programming language where interactive programming is basically the default way of working, and basically the same as normal programming in it, except the act of parsing is also the (in-memory) compilation at the same time, so it's running directly on the hardware instead of via interpretrer/vm/any of that overhead crap.
also then kinda open-source by definition.
and then to "only" write an OS in that, and voilá! a smalltalk-like environment with non-exotic, c-family syntax and actual native performance!
ahhh... <3
* a man can dream *2 -
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. -
How resource calculations for software services like code analysis, monitoring, etc are done:
Opening fridge, putting all the beer one can find in it.
Opening the necessary tools, e.g Excel, Accounting software, ....
Drinking the first beer.
Starting to aggregate the monthly costs - cause you can never trust the reports written by someone else...
First beer poof.
Looking at the monthly cost, adding columns "Intended use", "Actual usage pattern", "Usage factor"...
Opening next beer...
Usage factor is btw a factor of 0.1 ... 1.0 - to give an estimate how much the products feature are actually used, for further analysis if the invest is justified or not...
Oh. Another half bottle gone...
Filling in the columns...
Oh. Bottle empty and the next one toooooooooooooooo...
*burping*
*cracking finger joints*
Now let's get to the sad part...
Next worksheet, adding infrastructure costs...
Cost and description as columns.
Hehe. Column sounds like gollum.
Another beer...
Ugh. Need the paper reports, manually typing off things for stuff that was e.g. tax deductible.
Many beers die during this task. Poor little beers, dying for such an boring and mundane task...
SUM is a real useful function. I don't think I can add numbers anymore.
Now we can add another sheet.
Hehe. Sheet sounds like shit. And yes, everything in this file is shit.
Summing up costs from both sheets and including the cost factor from 1
... Beeeeeeeer Beeeeer beer we need more beer here... Beer beer beer...
Where was I. Oh yeah. Cost factorization total vs effective.
Why do I want to get even more drunk.
Oh yeah. Most software is completely underused and the costs aren't justified.
Let's add some colored highlighting ...
Uuuuh. ,Too much red. Better change the highlights.
Too much red.
More beer.
Don't give a fuck.
Hm.
Time for some whiskey.
What else is there to do....
Oh yeah.
Diagrams.
The bloody wankers from accounting need diagrams as numbers are too boring.
Not that everything in accounting is boring, no matter how much you paint colors on it... *sigh*
Hm. More whiskey...
Hehe. Whiskey rhymes with frisky.
Uff. Now just need to write mail. Mail mail mail....
"Copy paste the last mail from last month"
Hm.
Ah.
*sipping whiskey*
Spell check extension - to the rescue.
Thesaurus *burps*.
Let's change a few words here and there... Maybe another paragraph there.
Uh....
Trying to attach file...
*fucking mouse is pretty constantly crashing into empty beer bottles*
Done.
Damn.
Need to press send button.
*Creating mess on the desk by just randomly crashing the beer bottles*
Done.
*Pressing computers power button*
Mwahahahaha. No mouse needed.
*regretting to stand up too quickly, nearly barfing on the floor*
Couch ... Where Couch...
After hitting several doors, frames and other stuff, the glorious mission ended successfully with a most graciously executed gut buster on the couch.
(Regretting next morning to have emptied two 6 packs and a few glasses of whiskey) -
Often when i see the annoying as hell t debug exceptionless let’s just bomb entirely but blazing fastness of c and c++ I feel like a nettard
I use c# for its immutable strings clean syntax and beautiful class markers that are redundant compared to c++ but ensure you tell after adding 1000 methods and total lack of all special characters to indicate reference and derreference and pretty lambda syntax... sure it’s lib poor but I get shit done goddamn it and can read my own code later
So why do I feel empty inside every time i run a ./configure and make under Linux like I’m missing some secret party where neat things are being done and want to sob like I do now
I am not a dotnettard even though 5.0 is an abomination in the eyes of man and god ! Even though Microsoft cooks up overcomplex framework technologies that make a wonderful language underused and make us all look like idiots that they then abandon into the scrap heap! We can’t help Linux users haven’t discovered how much nicer c# is and decided to implement it on their own and port their horrible undocumented ansi c bullshit can we ???? Oh god I feel
So hollow inside and betrayed ! Curse
You gates curse youuuu! Curse you for metro direct3d xna wpf then false promises of core ! May you have a special place in hell reserved for you and your cheap wallpaper shifting monitor paintings and a pool speaker that playeth not but bee jees and ac dc forever and ever amen !
Speaking of which do any c/c++ ides have anything that even begins to rival intellisense on Linux and don’t use some weird ass build system
Like cmake as their default ?
Oh sweet memories of time a while back when I already wrote this and still wasn’t getting then tail I deserved
Again4 -
What are exemples of underused data types that you wish we used more often?
Using Set or Map or WeakMap instead of the normal Object+handmade dirty functions (JavaScript), for exemples?1
Top Tags
Weekly Rant
View