anarazel

49 ms·

571.
by anarazel 7y ago
The US political proceed being so fucked up that it can't react in a meaningful and halfway coordinated way is structural. That's far from the whole problem, but it's part.
572.
by anarazel 7y ago
Looks like it's also on increases, yes: https://www.sec.gov/oiea/investor-alerts-bulletins/investor-...
573.
by anarazel 7y ago
I am working on optional io_uring support for PostgreSQL. Hope to get that stable for PG 14 (PG 13 enters feature freeze soon).
574.
by anarazel 7y ago
There is, and has been for quite a while an option to change that. Look for reentrant / pure.
575.
by anarazel 7y ago
I think my stance on that is https://news.ycombinator.com/item?id=22459095
576.
by anarazel 7y ago
I don't think it every was a promising approach to let the compiler alone do that work. For one, it means that you basically have to recompile for every microarch, and the compiler needs all those details. But even then, the compiler c
577.
by anarazel 7y ago
What about it is beautiful? There idea of VLIW?
578.
by anarazel 7y ago
I don't think that was inevitable. If the Bush admin hadn't fucked up on such a grandiose scale after the invasion the result could have looked quite different (still surely far from pretty).
579.
by anarazel 7y ago
It doesn't seem surprising that it's portrayed that way, given that it's from a (former) high up member of the military. It's not like failures tend to be popular. There's plenty accounts describing it largely as a
580.
by anarazel 7y ago
One of the recurring complains / concerns of Pakistan is/was that Afghanistan post invasion was cooperating too much with India. From what I understand (which is certainly very limited) that was more fear than reality. But from th
581.
by anarazel 7y ago
There were over 100k US troops deployed to Afghanistan at the peak (part of 2010-2011). Between 2009 and 2013 it was over 50k. The Soviet Union deployed over 115k for several years. The amount of fortune spent was a serious accelerator for
582.
by anarazel 7y ago
Re bush: They also refused to negotiate / reintegrate the Taliban, in particular early on when they were weak (yes, there's a lot of parallels to Iraq). Once the Iraq war started there was very little attention paid to Afghanistan
583.
by anarazel 7y ago
A lot of the US problems around covid handling seem to be coordination related. There used to be coordinating personnel specifically for that at the NSC. But that has been slashed over the last few years. That kind of short sightedness driv
584.
by anarazel 7y ago
> Thanks for the numbers. Can you run it for -M extended as well? That would be the case when using unnamed prepared statement. Those are worse, due to the increased number of protocol messages (parse/bind/execute vs exec). tps
585.
by anarazel 7y ago
Is the 2/3 for caching using part of each NVMe, or is there one dedicated for the rest? Mostly wondering about spreading writes more evenly, rather than peak throughout.
586.
by anarazel 7y ago
It's pretty easy do completely misjudge the real world bottlenecks when still in the design phase. Can often lead to adding complexity to the design to alleviate a guessed performance problem. Making it harder to then later fix the tea
587.
by anarazel 7y ago
> Postgres planner is very fast, so there will be no noticeable performance difference by having to plan the same query again and again. This is unlike, say MSSQL, where a semi-complex query may take more than 0.5s to parse/plan, th
588.
by anarazel 7y ago
In fact, I'd say it's extremely common for database workloads to be compute bound on modern hardware. For one, storage bandwidth has increased massively (my laptop's NVMe drives can do 2 x 3.2GB/s reads, you can get quit
589.
by anarazel 7y ago
Yea, we really need to improve the handling of those cases. I think there's four (was three) major angles: 1) I'd hoped to get caching for JITed queries into 13 (or at least the major prerequisite), but that looks like it might mi
590.
by anarazel 7y ago
> Production MySQL and Postgres master databases that I administer typically have uptime loads of 0.1 to 0.2 on Linux, processing 5,000 qps. I have worked with plenty production postgres instances that have much much higher loads. As in
591.
by anarazel 7y ago
One TB of memory is actually quite expensive. And uses a fair bit of power.
592.
by anarazel 7y ago
Freeing memory also isn't free - the bookkeeping necessary, both at allocation and at free time and potentially also for the allocating code has costs. In postgres memory contexts are used extensively to manage allocations. And in quit
593.
by anarazel 7y ago
I think this long predates github. Sure, it might be less of an issue if it's just a one off code drop - but there've been maintained open source projects LONG before github. To me this is much more about reaching a (sub-) project
594.
by anarazel 7y ago
Because 4/8chan always stay there, and never venture out, right?
595.
by anarazel 7y ago
I'd hope that their data is all encrypted at rest. Compared to the bandwidth of spinning disks, the cost of doing hardware assisted AES isn't big.
596.
by anarazel 7y ago
You can detect whether a target is a plain table or not. And use a different path if not. And independent of using SQL or not: Executing them separately out of process kills you via latency, even if you're on the same machine. The cont
597.
by anarazel 7y ago
> Instead of worker processes and low-level row ingression, pgcat uses sql template to apply high-level sql commands, so it could make largest compatibility on target table type. It is written in golang and runs in separate process. Doin
598.
by anarazel 7y ago
I got a box with several items I ordered (damaged), but also a bunch of metal piping (bent), two lsat training books (damaged), and a bunch of other crap I can't remember. IIRC the stuff I ordered was a filing cabinet or something o
599.
by anarazel 7y ago
> 1) You don't know if a given linux kernel/other software will work unless you test it ... for each future version Huh? Sure, some software may break, but there's more than enough AMD out there to make sure that linux and
600.
by anarazel 7y ago
It's needed before the actual contents of the home directory are available (i.e. mounted), if I understand correctly. Nor is it actually user-modifiable: > Since the user record is cryptographically signed the user cannot make mod

More ›