101010.pl is one of the many independent Mastodon servers you can use to participate in the fediverse.
101010.pl czyli najstarszy polski serwer Mastodon. Posiadamy wpisy do 2048 znaków.

Server stats:

481
active users

#srm

0 posts0 participants0 posts today
Replied in thread

@Sustainable2050
And this issue was first looked at 20 years ago or more. Researchers were looking at how shell forming marine life was impacted by the increased acidity accompanying the increased CO2 concentration.

And for some reason the people pushing #SolarRadiationModofication (#SRM) keep forgetting that increased biosphere CO2 has adverse impacts beyond warming. The impacts on ecosystems cannot be ignored.

#ClimateCrisis is complex and technical wizardry will not solve it

**Setting up a Synology router and wifi mesh – was it a mistake?**

TL;DR

  • The initial problem was that ISP’s modem/router froze every day.
  • Synology is a good router with a responsive support. But –
  • The problem with Synology was responsivness of my selfhosted webpages via browser/domain name.
  • The solution was an obscure setting that ‘improves compatibility of roaming of wifi clients’
  • I lost 1 week of time to deal with it.
  • Nevertheless, it pays off to have own router. Innobox was s**t.

The problem:

My ISP’s (Telekom SI) modem/router (#Innobox G92) was having serious problems and ISP didn’t (want/know) how to fix it. It froze at least one time every day. It showed internet is connected, but it refused to transfer the data. Of course they (ISP) didn’t admit there is anything wrong with it. But I had logs in my HomeAssistant that showed exactly when the data transfer stopped.

They even replaced it and the new one started to behave exactly like the old one after only a week of use. They also blamed me – they said I should hire a ‘computer specialist’ to check my network.

I have only about 30 network devices.

So I decided (too late) that I want a router that I can control.

The next issue was some of my wifi gadgets didn’t have good wifi connection (e. g. in the basement). I used an old Linksys WRT54G as a separate access point, but I wanted a central management of all my gadgets.

Synology – the solution to ISP’s modem/router connectivity, which itself became a problem

I researched a bit and asked around, which routers are good nowadays. I heard about Ubiquity (and almost decided for it), Asus, TP-Link, OpenWRT (compatible) and Mikrotik. And Synology.

Finally I decided for a Synology. I have their NAS for 13 years, it has a nice UI, hasn’t failed yet and that was a deciding factor.

I went with their top offer: Synology RT6600ax router + WRX560 access point.

Installation of the router was dead easy. I just turned it on, connected to its wifi, created a user, connected WAN port to modem’s LAN port, called Telekom to put the modem in the bridge mode, entered PPPoE user/pass and voila, it worked.

SRM (the management UI) is really nice. I can see traffic by clients/protocols/apps/…

Wifi mesh access points and traffic:

Issues with internal web pages load times … was it DNS?

I tested the network a bit, added WRX560 as a mesh access point. And then the issues started.

Internal web pages accessed via browser / domain names were slow or there was a timeout when loading.

It looks like the network doesn’t handle http/s request to internal web pages well AFTER adding wrx560 and creating a mesh.

Before adding WRX, the access to internal web pages was quick. Even before, when I was using only my ISP modem/router, everything was quick.

Firmware: SRM 1.3.1-9346 Update 12

Symptoms:

When I tried to load a webpage that is on my webserver in my LAN via my domain (e. g. this page, https://blog.rozman.info and some others), it took 1-30 seconds for a page to load (or there was a timeout). Especially if I reloaded the page in a sequence or clicked links on the same page in short time.

Setup:

GPON –> Modem (bridge mode) –> rt6600ax (router) –> wrx560 (ap)

I have static IPv4 (and IPv6). My web domain points to the router static ip –> port forwarding –> reverse proxy (in my LAN) –> web server (in my LAN).

If I connected wirelessly or wired directly to rt6600ax, there was no timeout or delay. If I accessed it from external network (e. g. via mobile data), no timeout or delay.

If I connected wirelessly or wired to the access point WRX, there WAS a delay or timeout. Even if I put wrx on the last place in lan (after a dumb switch) and connected to the same switch, there was a delay. It got worse if I clicked refresh on a web page quickly several times in a row.

I ran countless tests without success.

Everything else worked fast and ok. Speedtest showed 300/100 Mbs, ping was 2-5ms, traceroute was 3ms to my modem, dig was fine, nslookup was fine).

Response of the webpages via IP – also no problem – loaded quickly, under 100ms.

But when I accessed it via domain name, it stuttered. Browser / Developer mode / Network / Timings showed:

If the page loads quickly via IP and stutters via domain name … it must be DNS, right?

Then I wrote a little curl script that measures the response time to exclude browser issues:

curl -w "\nDNS Lookup: %{time_namelookup}s\nConnect: %{time_connect}s\nStart Transfer: %{time_starttransfer}s\nTotal: %{time_total}s\n" -o /dev/null -s https://blog.rozman.info

that returns (when it’s ok):

DNS Lookup: 0.005905s
Connect: 0.007009s
Start Transfer: 0.838036s
Total: 0.838527s

when it choked, it returned:

DNS Lookup: 0.008007s
Connect: 1.009419s //or 2, 3, 10 seconds
Start Transfer: 1.952299s //or 2, 3, 10 seconds
Total: 1.952632s

when there wasa timeout, it returned:

DNS Lookup: 0.006191s
Connect: 0.000000s //timeout
Start Transfer: 0.000000s //timeout
Total: 21.034533s

DNS lookup looks quick, but after that (waiting and connecting) it chokes. 

It felt like that wrx560 (or mesh setup) added some kind of recursive loop to the (DNS?) request. NAT Loopback? I don’t know, I’m not a networking professional.

Failed attempts

I was desperate and tried many things without success:

  • updated both devices
  • disconnected AP, reset, played with its settings, but when I added it back to the mesh, it reset to its factory state and AP mode anyways. After adding it to the mesh, it worked ok for a few minutes, then the issues returned. Repeated several times.
  • Tried to access AP, no success. I couldn’t figure it out how to access it. It’s not in the client list. I even connected it to the router via LAN instead of WAN port. I could see it (MAC address, no IP), but still couldn’t connect to it.
  • Checked my reverse proxy (nginx), optimized some buffering and caching
  • Enabled caching mods (mod_expire) on my webservers (apache)
  • turned off and on almost all settings I could find it in the router software.
  • Cleared the DNS cache countless times (ipconfig /flushdns in cmd and about:networking#dns –> clear DNS cache in Firefox)
  • I even asked Mastodon community for help 🙂
  • In parallel, I contacted Synology’s support. They were helpful and response times were quick (24hrs).
    • Firstly, they sent me some patches and I installed it on the router. No success.
    • Then they ‘adjusted something’ remotely. No success.
    • More remote ‘adjustments’.
    • Lastly, they ‘asked someone’ and gave me the final solution. But in the meantime, I found the workaround solution:

Temporary workaround solution – split DNS

I temporary solved slow response/load times by adding my web site domain names to my local DNS records on my Pihole (‘split DNS’). I wasn’t very happy, because this is no real solution, only a bad workaround. But it worked.

I waited some more for the support to fix it.

At this moment I was sure I didn’t fuck it up. There must be something in the mesh setup that disturbes connection response.

And finally:

After a week of a chat with the support, they suggested to check out some obscure setting hidden deep in the menus and turn it off:

Control panel / System / SRM settings / Enhance the roaming compatibility of clients between Wi-Fi systems (turn it off).

AND IT WORKED!

Immediately after unchecking this setting, the my web pages became responsive. Whoah.

At the end, all good.

But then I started to play with IPv6…

till the next time!

Disclaimer

The links to the products (or mentioning them) are not affiliate links and I don’t receive any compensation for linking.

Hashtags: #synology #srm #homelab #mesh #selfhosting #dns #network

https://blog.rozman.info/setting-up-a-synology-router-and-wifi-mesh-was-it-a-mistake/

#dns#homelab#mesh

Today I am participating in a NSF Workshop on #Climate Interventions using #SRM #SolarRadiationManagement

They are pairing up climate scientists like me with #SBE #SocialBehavioralEconomic researchers

To try to figure out the ethics, governance, etc. of #geoengineering our way out of #ClimateChange

new.nsf.gov/events/exploring-e

Here is what I said previously about doing the fake-volcano method:

fediscience.org/@atthenius/109

corriere.it by Paola Cacace
Campania tra le stelle: l’aerospazio tra presente e futuro

Il 26 settembre webinar del Corriere del Mezzogiorno con Luigi Carrino (Dac), Massimo Deandreis (Srm), Catello Di Capua (Ecosistemi), Pietro Persico (Ma Group), Giuseppe Sernicola (Leonardo)

Translated:
Campania among the stars: aerospace between present and future.

On September 26th, a webinar by Corriere del Mezzogiorno featuring Luigi Carrino (DAC), Massimo Deandreis (SRM), Catello Di Capua (Ecosystems), Pietro Persico (MA Group), and Giuseppe Sernicola (Leonardo).

#LuigiCarrino #DAC #MassimoDeandreis #SRM #CatelloDiCapua #PietroPersico #MAGroup #GiuseppeSernicola #Leonardo #Italy #Campania
napoli.corriere.it/notizie/eco

Corriere della Sera · Campania tra le stelle: l’aerospazio tra presente e futuroBy Paola Cacace

The same old Making Sunsets dudes shooting SO2 balloons off in the news.

cbsnews.com/video/california-s

I remain concerned about #SRM #SolarRadiationManagement #Geoengineering as a way to counteract #ClimateChange

fediscience.org/@atthenius/109

Fortunately, NSF is convening several panels next month on the “Ethics and Societal Interactions of Climate Intervention”. I’ll be participating w other experts.

Hopefully the science community will be bringing (finding a path to) clarity and reasonableness soon

Replied in thread

@Snoro bullshit: avec du solar radiation management AU SOL, il n'y a aucun effet sur le photovoltaïque (ni sur le vent mais je m'avance).

Et nous n'avons PAS LE CHOIX car le SRM dû aux aérosols de provenance fossile va disparaître lorsque les émissions cesseront.

meer.org/about

www.meer.orgAbout MEER | Restoring Earth's Energy BalanceDiscover how MEER restores Earth's energy balance. Meticulous research, collaboration, and sustainable solutions for a resilient future.

Someone is going to dim the sun, and it will be soon, climate.benjames.io/someone-is.

The fact:
- “100 planes injecting sulfur particles into the stratosphere would dim the sun by about 1%, and cool the earth by about 1°C”

The reality:
- We don’t know if a cascade reaction can happen, and if it will, it can be absolutely dramatic.
- It is most costly than the article suggests.
- We need to regulate/forbid this.

Ben James · Someone is going to dim the sun, and it will be soon.People are starting to take geoengineering seriously.

This article on #SRM (solar radiation management) is a model of circumspection in the best and most formal sense.

Deep enough for most of us while also highly readable. Salute to author.

SRM is undesirable in the same general way as is using "fix a flat" on a car tire with a puncture; there are several downsides, starting with "temporary" and "messy." Equally, sometimes unavoidable.

#SolarRadiationManagement
#MarineCloudBrightening
#StratosphericAerosolInjection
theregister.com/2023/12/30/sol

The Register · Scientists mull Solar Radiation Management – a potential climate-change stop-gapBy Rik Myslewski

Despite it's title "It's Time to Engineer the Sky" this piece in SciAm does a decent job of laying out the risks of geoengineering using stratospheric aerosols as well as the arguments people use to promote research into it.

Original title "A Stratospheric Gamble" was a better one though.

Bottom line: SRM has major risks but also quite possibly deployed in future. I'd rather not leave research up to carbon offsetters

#GeoEngineering #ClimateChange #SRM

scientificamerican.com/article

Scientific AmericanIt’s Time to Engineer the SkyGlobal warming is so rampant that some scientists say we should begin altering the stratosphere to block incoming sunlight, even if it jeopardizes rain and crops

Passwords and Peanut Butter

Wanna buy some *really* expensive peanut butter?

For a modest investment of $1,069.00, you too can acquire a tasty 3-pak of NIST Standard Reference Peanut Butter.

Coming in at approximately $59/ounce, that’s a bit pricier than the $.20/ounce you’d spend at the grocery store. But in this case, you’re buying more than the product in the jar. You’re buying “truth.”

NIST, the National Institute of Standards and Technology, is well-known as the government agency which sets password policies. NIST is the reason most websites began asking users to create passwords which contain:
☑️​ Upper-case letters
☑️​ Lower-case letters
☑️​ Numbers and
☑️​ Special characters

NIST is also the reason most websites now ask for passwords longer than 8 characters, and why websites *don’t* ask us to choose from a list of goofy questions like “Where did you attend high school?” or “What’s your favorite food?”

In the world of cybersecurity, NIST is almost always thought of as the good guys.

In addition to defining best practices and providing advice related to technology, NIST also produces 1,300 Standard Reference Materials, or SRMs. Thirty of these are food-related.

These highly-analyzed samples are used by manufacturers all over the world to calibrate their testing equipment. Lab techs know if their results agree with the NIST standard sample, their tests are reliable.

The high cost of these reference materials — including peanut butter — is really all about the extensive process NIST used to accurately measure the ingredients, and all the scientific expertise that went into verifying the sample’s chemical and physical properties.

NIST refers to their Standard Reference Materials as “Truth In A Jar.”

Other fun facts about these reference samples:
➡️​ NIST sells $20MM per year
➡️​ From a 20,000 square foot warehouse
➡️​ The most popular SRM is a break-away product called “Charpies”
➡️​ NIST even sells standard cigarettes for flammability testing

The newest Standard Reference Samples are:
➡️​ Typical Human Diet
➡️​ Human Sludge (ugh!)
➡️​ Live Hamster Ovary Cells for production of human monoclonal antibodies

So NIST isn’t only all about passwords. They also help lab techs find nasty impurities in peanut butter, like aflatoxins, that would do us a world of hurt if not for the vigilant folks at NIST.

Here's a recent behind-the-scenes video from Veritasium as they tour the NIST SRM warehouse:

youtube.com/watch?v=esQyYGezS7

#NIST
#SRM
#PBJ
#PeanutButter