I'm considering switching from vanilla arch and so I've been using Garuda on an external SSD occasionally.
My machine is not slow by any standard, but even with a light and optimized arch setup, webpages load slowly. Whereas in Garuda, they load relatively instantly—this is the performance I would expect on modern hardware and high speed internet. I could alternatively ask why pages load so slow on my regular system, but I can't think of any reasons.
Additionally, where I am the wifi is shared like at a university. I've had to resort to using a VPN to try to get even remotely decent loading speeds and download speeds. But when installing Garuda or doing package updates, I get the full 20+mb/s—all of this with no VPN.
The weird thing is, page load speeds are not necessarily purely based on internet speeds. The page data may load relatively instantly and then take a long time to render—but I'm using the same Firefox browser with the same extensions.
The use of the profile sync daemon can significantly improve browser speed in some cases.
The other thing that comes to mind is if your Arch machine is using swap during these browser sessions, that will slow things down a lot. Garuda uses zram by default, which will spare you the performance hit.
Speaking of browser speeds, if you use Firefox (I use Librewolf a more secure version) and Dark Reader, which is a popular add-on to make all web pages ....dark, it will slight slow down browser speeds. It is horrific on Android devices. I couldn't figure out why it would take a whole minute to load my pages, then I did a search and found out its a known issue with Firefox and Dark Reader, just fyi.
Yeah, me either. It is not so noticeable on the PC, especially if you have a lot of ram, etc., but man, Android it sure is, just try deactivating it and doing a search. The results appear instantly. I had to let it go on Android. BTW I use Fennec (Firefox secure) on Android.
Interestingly, Archinstall defaults to ZRAM usage.
@EmeraldEnigma, if straight Arch is slow there is something wrong with your setup. Even using vanilla Arch+KDE+VPN, there is no apparent issues with speediness that I have encountered. Garuda does have a few optional settings that can impact real and/or perceived speed but those aren't very difficult to setup yourself in Arch or any other Linux distribution. However, being able to do so with just a click of a mouse button in Garuda is most certainly a time-saving measure.
How did you install Arch? What do you use for default settings? Those have a great bearing on load/operating speed. That, and the fact you are at the mercy of your University-like shared Wifi.
You're still using the same network in either situation.
I use archinstall and it usually says something about zram but I don't remember where or why (maybe when I enable swap).
I wonder if garuda installs with official arch mirrors, because there's no question that consistently garuda gets what I assume are max speeds when downloading things, whereas if I'm using arch without a VPN (e.g. during installation whether over wifi or ethernet, or even during the initial configuration stages) the packages download SUUUUUUUUUPER slow.
What I was noticing with gardua was before I had configured anything, so no vpn, no dark reader, search results showed instantly, reddit pages loaded instantly, and so forth.
I'm not sure if the package speeds and the internet speeds are related.
But how does that explain why the packages download fast (on vanilla arch) once I enable the VPN? (e.g. on the base system after I've done some initial configuration, or when installing arch in a VM where it's using the internet from the host which has VPN enabled)
Yes. The Chaotic mirrors could be considered not official Arch mirrors, but will not have any impact on what you are describing.
If you are getting faster speeds when connected to your VPN, a likely explanation is the ISP or whoever is administering your network is traffic shaping with QoS or similar. On a network with pooled resources like you describe, this is very common. An admin can assign a low QoS value to stuff like downloads to preserve bandwidth for things like video calls or whatever.
An interesting test would be to host a video call both on and off the VPN, to see which gets a better performance. If your results are the reverse of the download situation (i.e. off the VPN is better), it is likely QoS is the culprit.
Traffic inside a VPN cannot properly participate in QoS because the encrypted tunnel makes it impossible to determine what kind of traffic it is. VPN traffic gets assigned a “blanket” QoS value, which in your case could potentially be a higher bandwidth than whatever bandwidth downloads are getting after being throttled for traffic shaping.
Regarding the QoS I'm 100% certain that's the case. This was obvious within minutes after getting a VPN. Still...
[USER] What does this mean?
The secret is that our RFC 2549 implementation is faster... the avian makes the difference. :slight_smile:
[ChatGPT] This sentence seems to be referring to the speed of an implementation of RFC 2549, which is a specification for transmitting compressed audio over IP networks. The author of the sentence is claiming that their implementation of RFC 2549 is faster than other implementations, and that the reason for this is an "avian" (which means "bird") - possibly a humorous or obscure way of referring to a software or hardware component.
Without more context, it's difficult to say for certain what the author means by "avian." It's possible that it's a reference to a specific technology or hardware that they're using to speed up their implementation, or it could simply be a joke or play on words.
I got the bird part. The RFC 2549 thing must be an inside joke?
Traffic shaping makes a lot of sense. That's one reason I now subscribe to VPN--to avoid the traffic shaping my ISP had been threatening--not that I didn't deserve it.
On the serious side, none of the large business networks I've ever used resorted to traffic shaping for any purpose. But then, it's been a few years.
Archinstall uses the Reflector-chosen mirrors when you boot up an Arch LiveISO. (Garuda uses a different utility.) Here’s how a dev explained a method of selecting a specific mirror when using Archinstall.
I also edit /etc/pacman.conf after booting into an Arch LiveISO prior to running Archinstall.
lmao, it does try really hard, but occasionally it hallucinates (the word for when LLMs just invent some convincing nonsense that fits the rest of the context)
However, newbs should not be sleeping on ChatGPT. As someone who started getting serious with linux about 1 year ago, and started programming infrequently within the last couple years, it is MIND BLOWING how helpful it is.
It will literally write any linux command you need, and in the reverse it will also interpret any linux command you don't perfectly understand (see pic).
It can write scripts, and though it might make a few mistakes it can debug code which means it can fix both my scripts and its own scripts.
It helps explain obscure linux stuff and packages.
Overall while it's not 100% flawless, it has been bang on for about 95% of my use cases, and has ultimately saved me countless hours of frustration and immeasurable stress from having to go on linux forums and have barely-human jerks tell me that if I can't follow the arch wiki then I shouldn't use linux.
It's like having a big bro who knows linux, always there ready to help with any issue instead of telling you to "haha read the manual lol" .
Ya I've looked into trying to use better mirrors with archinstall before. That thread about covers it: you try, and realize it doesn't work. The final comment seems to explain how to make it work, but it's over my head, and anyways it looks like it uses a step from his previous comment where it said it can only use one mirror.
.