nezirus

I'm mostly interested in numbers under 100W, maybe even 65W, since I plan to down-clock my (future) Zen 4 machine.

AMD is killing it with efficiency there.

You don't need podman-compose, docker-compose is perfectly viable with podman. They've even fixed some long standing bugs/problems lately, so docker-compose v2 can also be used with podman [1].

[1] https://podman.io/releases/2022/05/09/podman-release-v4.1.0....

Nice story! It also matches my experience with many open source projects. It is not enough to solve a technical problem, you must convince other people to accept the solution.

Having a nice maintainer to guide you through the process is really encouraging and brings more nice and bright people to the comunity. Kudos

If you are concerned about RAM usage, there are some trade-offs available (slightly lower compression ratio), for more effective resource usage.

As used by HAProxy for gzip filter, from the Will Tarreau, creator of HAProxy: https://github.com/wtarreau/libslz

They were first to push crun[1], in place of runc, that is your C code in the stack :-)

[1] https://github.com/containers/crun

Aside from the kill count calculations, the hardest thing for me here would be actually to be in the "position of power". Who are we to decide such things, even if we have some info (or think we have).

It is actually the oposite on Linux. You compile on older distro (e.g. CentOS 6), and it will work on later versions, (CentOS 7).

https://developers.redhat.com/blog/2019/08/01/how-the-gnu-c-...

Actually, it supports docker-compose proper (v1 and v2, even if there are some bugs for v2)

Podman also works fine rootless and with cgroups2, double win.

Sway has multiseat, maybe it would be a match for your use-case, check manual for sway-input.

Maybe this can help https://github.com/awslabs/pgbouncer-rr-patch

It is a pgbouncer fork/patch with query rewriting support.

You are correct, the proper way to do deletions in ClickHouse is to use partitions, and drop partitions. That is probably good enough for most analytical use cases, but YMMV.

The article is really poor and not original, doesn't link to haproxy or jfrog web pages or haproxy git repo.

Please link the original jfrog blog post with in depth analysis:

https://jfrog.com/blog/critical-vulnerability-in-haproxy-cve...

I kindly disagree. It may appear that chess is a memoization game, but when or if you want to become a master of the game , you'll notice that rote learning will not take you there.

You do need to understand phases of the game, set up strategy, estimate your position, know when to attack or to defend, and have confidence in your technique.

Mind you, that is only for the plain master title (FIDE or even a national master). Grandmasters are a whole new level in comparison, aliens.

You just do raw byte copies from sample DB, no SQL or "inserts" or anything similar. Imagine test database consists of 3 parts (all raw bytes)

### PROLOGUE ### Sample row ### EPILOGUE

You copy & write prologue, write 1B sample raws (can optimize this at will, large writes, etc)

Copy & write epilogue and fsync the data. You probably need to modify some metadata, but that should be a few writes at most.

That should be as good as it gets, providing your IO is optimal.

Dunno, we're probably fucked up anyway. Our education system is destroyed, splintered into many small pieces, low quality, that is the worst thing for the future. There are simply too many layers of government, half of the country considers itself part of Serbia, the other part is busy crying wolf (while stuffing money into pockets). Teritorial rearangements wont make people more educated, happier, wealthier...

Even welcoming Bosnia into EU wouldn't help, that would be just a free pass to relocate for work, and a shot of a decent life (that is not easy to achieve in a foreign country, working jobs locals would not take)

For starters, don't reward war crimes. Organize swift trials. They've only recently finished trials of high ranking generals and officials, there are thousands lower ranked ones which will never feel the hand of justice.

Jasenovac happenned during WWII, one of the most despicable events in history of the Croatia https://en.m.wikipedia.org/wiki/Jasenovac_concentration_camp

However, the memory of Jasenovac often serves as justification for new atrocities commited to revenge the victims (50 years later!)

Serbia was bombarded for crimes on Kosovo, probably to stop Kosovo from becoming next Bosnia...

I can't speak for other ex Yugoslav republics, but Bosnia and Herzegovina is in permanent state od limbo, for 25 years, and probably for the next 25. Corruption is rampant, nationalistic retoric abundant...

One of the main reasons we are a failed state is Dayton peace agreement. It awarded war and genocide. Just imagine post WWII Germany, splitted in two parts, one ruled by ex facists, still denying holocaust, the other part ruled by self serving, corrupt, thieves and bastards (can't say they are better than facists, because all of them got into power and got rich during the war)

The whole region will have half of the population (from year 1991), in just a few decades. Young people/families will never return...

Such a tragedy, to execute a teenager for such "crimes". It makes me sad, and angry... Not only they kill children in Yemen, but also their own citizens.

What can we do about this, maybe not much, these tyrants are supported by their friends in the West, but we can surely try. Pressure politicians to stop supporting them, selling them weapons, allowing their dirty money into the financial system.

Maybe even smaller steps could make an impact, like what if Muslims stop going to Hajj as long as these monsters are in powers. They earn big money from that.

Deleting whole partitions is generally useful strategy. It's like the difference between single inserts and batch inserts (often huge performance difference, and much lower IO)

Since you mentioned Cassandra and TTL, I'll mention ClickHouse, very nice TTL options, splitting into smaller partitions and using "ttl_only_drop_parts=1" has prove itself in the production with big data ingestion rates.

Last, but not the least, I almost always prefer Postgres for data storage needs, one can trust it to be safe and fast enough. Only some specific situations warrant other solutions, but it's a long way until that point (if ever), and better not optimize too early.

pgbouncer has PAUSE comnand, which can be used for seamless restarts/failovers or similar

Postgres too https://www.postgresql.org/docs/current/file-fdw.html

Yep, my experience too. Critics of Israel get called antisemitic too easily, moderated down here on HN too. On the other hand, people who defend atrocities committed by IDF get easy pass, openly defending stuff like: "the IDF is really gentle in killing civilians (collateral damage)", they nicely warn Palestinians "we're going to destroy your house, so please leave in 5 minutes", or "be silent while we bulldoze your house", etc.

Back on topic, the blog post title is definitely offensive, you can't label people like that. Maybe it was a bad tongue-in-cheek, thinking about it, I don't think you can even say Israelis or citizens of Israel, since not all people have the same political views or support the same solution of the "Palestine problem".

Yeah, the whole idea of fasting is moderation. So if you over-eat yourself, that defeats the purpose.

We (modern people) probably eat to much, so some kind of self discipline, either fasting, or avoiding sugars should be beneficial.

The Java numbers (1 CPU) don't look so good to me, especially compared to P99 and memory consumption for Rust and C++

Results for 2 and 3 CPUs are strange too. So little scaling, what's the point...

Yeah, don't want to provide free advertising (they don't need it), but they should be as reliable as OVH for hosting servers in Europe.

I don't have enough details about the exact setup, but the server cost looks excessive* to me. Especially those Xeon Gold and Silver machines (are they using managed hosting or what). Almost all of them are in single OVH data center too, that's not very robust.

* Compared to a certain German hosting provider, where I can get AMD EPYC 7502P with 128GB RAM for about 100EUR. Again don't know the arrangement, but with some effort they could add more servers, lower the cost and add Geo-resiliency.

Yeah, I'd like to see that as well, something like to_string_lossy(), but really fast, so I can call it whenever I expect occasional broken UTF-8 string (and don't care about few replaced characters).

In addition to other complaints about this article (old, pretty shallow, testing methodology etc), one thing caught my eye:

Title: "How much faster is Redis at storing a blob of JSON compared to Postgres?" First sentence: "tl;dr; Redis is 16 times faster at reading these JSON blobs."

I know there are numbers for "writing" as well (20x), but this is just poor writing (pun intended).

What the hell HN, how come stuff like this bubbles up to the front page?