Welcome to
This week: Intel optimizes Linux multithreaded networking, data centers in space (again), and more DevOps toolmaker layoffs.
First up this week: Intel proposes a patch that smooths out a bottleneck in Linux networking. Contention for shared data structures across threads turned out to be sub-optimal.
Moore’s “law” implies we’ll continue getting more and more cores on a single chip. So anything that helps multi-threaded throughput should be a good thing.
Michael Larabel: Intel Continues With More Big-Time Optimizations To The Linux Kernel
“Sierra Forest”
Intel engineers have been sorting through a bottleneck within the Linux kernel networking code and discovered a performance issue around concurrency with the dst_entry data structure. … I couldn’t help but to try out the patches myself.
…
[I] saw striking performance improvements now when looking at the memcached performance. … Some massive speed-ups. … This optimization work also isn’t only limited to Intel hardware but should benefit other processors too.
…
It’s wonderful seeing all of the Linux kernel optimizations that continue to be pursued by Intel engineers as well as their many optimizations and enhancements throughout the rest of the stack too—from compilers to other key libraries. These … scalability optimizations will also become all the more important and beneficial with higher core count Sierra Forest processors next year.
ELI5? chaorace explains like we’re five:
Basically, processes that produce a lot of threads (i.e., 100+) suffer performance degradation. This impacted network performance by up to 5% on 1 Gb/s connections and showed up to 3.2x performance degradation in synthetic benchmarks.
Independent testing verifies these improvements in the Memcached benchmark. Simulating a more realistic workload using Cockroach DB shows a more modest performance improvement of ~5% in … ops/s.
And kozman doffs a cap:
Reminds me also of the BFQ [I/O scheduler] story: … A simple tweak that just required some effort.
…
All these little nuggets are just waiting for someone to come along and put in some sweat equity and all the Linux community gets a boost. So cool. I have mad respect for the code ninjas who are able to figure this complex stuff out.
A Floridian startup has found enough suckers raised sufficient funding to send a small data center to the lunar surface. The company claims that, later this decade, it’ll be able to offer DR and edge compute services.
If this idea sounds a bit “out there,” that’s probably because it is. And so is the moon.
Passant Rabie: Florida Startup Moves Closer to Building Data Centers on the Moon
“Hopes to deploy on the lunar surface by 2026”
Cloud computing startup Lonestar Data Holdings … raised $5 million in seed funding to … establish a viable platform for data storage and edge processing … on the lunar surface. … The company is now ready to launch a small data center box to the lunar surface later this year as part of Intuitive Machines’s second lunar mission, IM-2.
…
The lunar data centers will initially be geared towards remote data storage and disaster recovery, allowing companies to back up their data and store it on the Moon. In addition, the data centers could assist with both commercial and private ventures to the [moon].
…
The miniature data center weighs about 2 pounds … and has a capacity of 16 terabytes. [It] will draw power and communications from the lander, but the ones that will follow … will be standalone data centers that the company hopes to deploy on the lunar surface by 2026.
Cloud on the moon? Imagine the latency! The keyword here is “edge,” notes test321:
It seems what they want to do is to be the IT department of future permanent moon settlements. The ‘data center’ will store mission data and perform the computing they need (e.g., train AI models on acquired videos), rather than sending many terabytes back to earth, then back from earth to moon after processing.
…
Think of it as the old saying about the bandwidth of a [station wagon] full of tapes. If a yearly moon mission brings a dozen new 18 TB disks and brings them back full of data, it’s equivalent to 7.2 Mb/s bandwidth 24/7 over one year.
How could this possibly work in such a harsh environment? It can’t, as WRXforScience explains:
The Moon has no atmosphere, so even dust size impactors (micrometeorites) make it to the surface while traveling at thousands of miles per hour. … Additionally, the dust on the Moon is incredibly abrasive and electrostatically charged. There’s no wind to move it around but there are charge gradients from day/night transitions that kick up dust.
…
The lack of an atmosphere also makes heat dissipation tricky. Large radiators are usually needed to get rid of excess heat. There are some interesting options to use the lunar equivalent of geothermal systems to dump heat into the ground on the Moon (lunar heat pumps), but we’ve yet to try to build one.
So senttoschool offends Florida Man:
No offense to Florida but the state attracts a lot of grifters and opportunists. I see this as another one.
500 people are being kicked out of Atlassian. And it’s all in the name of “rebalancing skills.”
This comes on the heels of last month’s CVSS-9.4 authentication bug and customer-PII-is-safe-honest data leak. Customers of Jira, Confluence, Trello and Bitbucket are feeling increasingly antsy.
Kate Ainsworth: Atlassian to shed 5% of workers, or 500 jobs globally
“On March 10”
In a statement released to the US Securities and Exchange Commission … Atlassian founders and co-chief executives Mike Cannon-Brookes and Scott Farquhar said, … ”A month back, we reorganised our company to better reflect operating in a changing and difficult macroeconomic environment. … While it helped us streamline work, we need to go further in rebalancing the skills we require.”
…
Atlassian has headquarters in both Sydney and San Francisco, and employs personnel across 13 countries. … More than half of the employees impacted by the layoffs are located in North America. … Employees who work in the talent acquisition, project management, and the research and insights teams are the most affected.
…
Those impacted by the job cuts will finish with the company on March 10. … Atlassian is the latest tech company to lay off a portion of its workforce, with Google, Microsoft, Meta, Amazon and Twitter all shedding thousands of employees in recent months.
What went wrong? devhead suggests one or two issues—actually six:
Some of their problems are tied to their severe tech debt, weak leadership, poor vision of the future, lack of understanding of the present, all the while relying on 3rd parties to profit instead of them actually making a better product and providing … changes with any sort of velocity. Oh, and they killed off their main competitor [to] Trello—but sweet now my user is deleted and I have to [be an] Atlassian user now. :: sighs ::
…
Hope they give these employees a good exit. It’s the leadership’s own fault.
Meanwhile, the news has reached the Ace2 receptor: [You’re fired—Ed.]
I’m fine with it, as long as they fire the people responsible for Jira.
—Dr. Seuss
You have been reading The Long View by Richi Jennings. You can contact him at @RiCHi or tlv@richi.uk.
Redis is taking it in the chops, as both maintainers and customers move to the Valkey Redis fork.
GitLab Duo Chat is a natural language interface which helps generate code, create tests and access code summarizations.
Expect attacks on the open source software supply chain to accelerate, with attackers automating attacks in common open source software…
The emergence of low/no-code platforms is challenging traditional notions of coding expertise. Gone are the days when coding was an…
Datadog today published a State of DevSecOps report that finds 90% of Java services running in a production environment are…
Linux dodged a bullet. If the XZ exploit had gone undiscovered for only a few more weeks, millions of Linux…