HNHacker News
TopNewBestAskShowJobs

man8alexd

130 karma · joined September 21, 2025

submissionscomments
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
Maybe you should skip running a useless child process and just use PSI to monitor memory pressure.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
You can always tell the system not to kill your Photoshop (or whatever else) by setting the OOM Score Adjust. This mechanism has existed for almost two decades and systemd has supported it for over a decade.
man8alexd··on FreeBSD ate my RAM
You will be surprised by how inaccurate memory measurements are.
man8alexd··on FreeBSD ate my RAM
See this retrocomputing.SE question https://retrocomputing.stackexchange.com/questions/32492/ori...
man8alexd··on FreeBSD ate my RAM
FreeBSD didn’t have memory overcommit and instead used strict swap reservation - each allocated anonymous memory page was supposed to have a corresponding swap page. This required 2x RAM swap space, otherwise you would get “out of swap” when forking a large process. FreeBSD implemented memory overcommit around 2000.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
No, it doesn't handle failing allocations gracefully - https://unix.stackexchange.com/questions/797841/firefox-died...
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
Is Firefox low quality?
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
People like to reinvent things that they are not aware of. Original BSDs used to use strict swap reservation - every anonymous memory page had to have an associated swap page. You had to have the swap 2x of RAM to allow large processes to fork - otherwise you would get an "out of swap" error. FreeBSD implemented overcommit around 2000, I think version 4.x or 5.x.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
MariaDB recently implemented memory PSI monitoring but failed with that in a curious way and disabled it afterwards by default. The failure is that under memory pressure, they flushed the entire InnoDB buffer pool.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
The last paragraph:

> On the modern desktop, where programmers don't care about failing malloc(), disabling overcommit is shooting yourself in the foot. As you can observe, the memory allocations start failing long before the memory is exhausted.

man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
They do have sidecars like prometheus, node_exporter running alongside Postgres and they include them in their MemoryLimit calculations.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
There is some kind of illusion or myth that strict overcommit solves memory management issues.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
cgroups have nothing to do with overcommit and memory allocation. They limit actual memory usage for a specific program or group of programs. If this program tries to use more memory than the cgroup memory limit, the program gets OOM killed.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
Yes, many have tried to use strict overcommit on the desktop. It is a good footgun. https://unix.stackexchange.com/a/797888/1027
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
Mode 0 (Heuristic) is described incorrectly. All this complex heuristic was removed almost a decade ago. Currently, the kernel refuses a single allocation that exceeds the physical memory. That is all.

The article ignores the proper modern solution to prevent OOM killing of critical processes - OOM Score Adjust.

Tuning CommitLimit manually is an archaic, imprecise, and error-prone way to handle memory limits, only suitable for single-process workloads that can handle ENOMEM properly. It completely ignores dynamic file page cache memory allocation. You still can get OOM if you get unusually high file activity. On the other hand, under low file activity, it wastes memory on the same page cache, because it can't be reclaimed without memory pressure, and memory pressure can't be created because workload hits ENOMEM earlier. Don't use strict overcommit.

man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
Windows doesn't need to fork, and you can't fork a large process without overcommit.
man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
disabling overcommit is trading OOM for random program crashes due to the inability to handle ENOMEM. It also wastes a lot of system memory.

https://unix.stackexchange.com/questions/797835/disabling-ov...

man8alexd··on PostgreSQL and the OOM killer: Why we use strict memory overcommit
> The Linux Kernel OOM killer kills random things.

By default, the Linux kernel kills the largest process in the system (unless OOM adjust was applied).

man8alexd··on Ask HN: Who wants to be hired? (July 2026)
Location: Belgrade, Serbia (UTC+2)

Remote: yes (working exclusively remotely for the past 20 years)

Willing to relocate: yes

Technologies: AWS, Kubernetes, Nomad, Docker, podman, Hashicorp Consul, Vault, Packer, Terraform, Ansible, Chef, Puppet, Grafana, Loki, Prometheus, ELK stack, Jenkins, GitHub Actions, Flux, GitLab, Postfix, HAProxy, MySQL, MariaDB, PostgreSQL, Ruby, Python, bash, C

Résumé/CV: https://alexeydemidov.com/cv

Email: in the resume

GitHub: https://github.com/AlexeyDemidov

LinkedIn: https://www.linkedin.com/in/alekseydemidov/

ServerFault: https://serverfault.com/users/23022/alexd

Have been playing with Linux and FreeBSD for over two decades. Run an ISP engineering team for a decade, now building and maintaining cloud infras.

man8alexd··on Ask HN: Who wants to be hired? (May 2026)
Location: Tbilisi, Georgia (UTC+4)

Remote: yes (working exclusively remotely for the past 20 years)

Willing to relocate: yes

Technologies: AWS, Kubernetes, Nomad, Docker, podman, Hashicorp Consul, Vault, Packer, Terraform, Ansible, Chef, Puppet, Grafana, Loki, Prometheus, ELK stack, Jenkins, GitHub Actions, Flux, GitLab, Postfix, HAProxy, MySQL, MariaDB, PostgreSQL, Ruby, Python, bash, C

Résumé/CV: https://alexeydemidov.com/cv

Email: in the resume

GitHub: https://github.com/AlexeyDemidov

LinkedIn: https://www.linkedin.com/in/alekseydemidov/

man8alexd··on Chernobyl wildlife forty years on
This information is not discussed in INSAG-7. It is from trial testimonies sourced from the book by Nikolai Karpan (deputy chief engineer of Chernobyl NPP), who was present at the trial and made notes himself.
man8alexd··on Chernobyl wildlife forty years on
Internet at its best - explaining how it was in the USSR to a former soviet citizen.
man8alexd··on Chernobyl wildlife forty years on
For the rundown test, after the valve cut-off, it is irrelevant whether the reactor was shut down or not. The working plan for the test only specified that it was supposed to run before a planned maintenance period, so the shutdown was implied. During previous tests, the AZ-5 signal was wired to the valve cut-off signal and was sent automatically at the same moment. It is not clear why this changed in 1986, but the outcome would have been the same if AZ-5 had been pressed 35 seconds earlier.
man8alexd··on Chernobyl wildlife forty years on
Easy but boring. Realistically, the workers would be gathered in a hall and their immediate boss would give a speech: "The party and the government want you to serve the Motherland at this heroic moment and volunteer for a hard job. Whoever goes gets apartments ahead of the waiting list".
man8alexd··on Chernobyl wildlife forty years on
The AZ-5 button was pushed as normal shutdown procedure as the test had been completed, not as a reaction to some event.
man8alexd··on Chernobyl wildlife forty years on
I worked in the soviet nuclear industry (Sredmash) in the 1980s.

The dialogs and characters are completely unrealistic and made me cringe. Everyone looks overemotional and infantile.

The hierarchical interactions are comical - a minister would never go to talk to miners, he would just phone a subordinate and tell them to organize people, they don't need armed soldiers present to enforce something, it is not the Wild West. The authors have no clue about the soviet mentality and how soviet society operated.

man8alexd··on Chernobyl wildlife forty years on
"Verbal abuse" isn't a concept that existed in the Soviet Union. Giving or receiving instructions with as many "suka blyat" inserted between each word as possible wasn't abnormal.
man8alexd··on Chernobyl wildlife forty years on
Quote from INSAG-7 https://www-pub.iaea.org/MTCD/Publications/PDF/Pub913e_web.p...

> neither the reactor power nor the other parameters (pressure and water level in the steam separator drums, coolant and feedwater flow rates, etc.) required any intervention by the personnel or by the engineered safety features from the beginning of the tests until the EPS-5 button was pressed. The Commission did not detect any events or dynamic processes, such as hidden reactor runaway, which could have been the event which initiated the accident. “

man8alexd··on Chernobyl wildlife forty years on
The TV show pretends to be historically accurate, and many people believe that it is true. I would suspect that the majority of people have no other sources of information about Chernobyl other than this TV show.
man8alexd··on Chernobyl wildlife forty years on
Higginbotham uses Medvedev's book as a source. Medvedev worked in the Ministry of Energy and he was their special representative in Chernobyl after the incident. His task was to cover the asses of the ministry and the reactor designers, so this book invented a lot of "facts" to put the blame on the operators, Dyatlov and Fomin.
← PreviousPage 2 of 6Next →