Performance concerns on "fe.disroot.org"/soapbox frontend. #177
Labels
No Label
administration
Akkoma
Android
Bare metal
bug
Communication
Community
Cryptpad
Discussion
Documentation
duplicate
enhancement
etherpad
Feature request
Feedback
finances
Fixed
forgejo
fun_project
Goal 2024
help wanted
Howto
In progress
🤔️ Investigate
ios
jitsi
lacre
Lacre Test
ldap
Lemmy
LibreTranslate
low prio
Lufi
macos
Mail
Merch
monitoring
movim
needs_refine
New Auth
Nextcloud
nice to have
on hold
proposal
question
Ready
refined
Roundcube
searX
spam-protection
Staging Server
Themes
TOR
Urgent!
Website
windows
wontfix
xmpp
Yearly Report
No Milestone
No project
No Assignees
3 Participants
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: Disroot/Disroot-Project#177
Loading…
Reference in New Issue
No description provided.
Delete Branch "%!s(<nil>)"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
I am noticing a definite performance lag with the new ActivityPub instance especially on initial load of page and sometimes first click to another area after the home page loads (like the local timeline, for example). Site seems reasonably responsive after this point so far.
Browser becomes slow to unresponsive for a couple of seconds. Mobile browser especially impacted (Android with FF Nightly). I am also seeing hesitation with both FF snap and deb package.
This user is still looking at what info to pass on to group, but creating issue for others to be pointed at. Looking at various FF web tools from the desktop version.
Possible example thread that impacted the browser greatly:
https://fe.disroot.org/@thegibson@hackers.town/posts/AEItU1ZeFTMxxPXlDs
Interesting. I thought this was caused by me having as usual over 200 tabs open and my laptop running on the edge of ram/cpu. I have now opened the link you posted and indeed the entire browser is affected by this. It will be good to check logs both server side and browser side and submit it upstream to soapbox, I have a feeling it something with the frontend. Thanks for pointing this one and submitting link we can use to reproduce.
As for mobile, I would not put too much effort into it as there is plenty of clients so using browser for AP services is just bad choice on mobile anyway and should be discouraged.
Yeah, I also have a tonne of tabs open and this is the first page that has killed the whole thing. I wanted to use some of the FF web developer tools, but if the whole browser is impacted; not sure what would work. Is there anything I can send your way to help?
WRT mobile front ends, I strongly feel that a responsive web front end should come first over apps/api for your typical user. Idealistically, I would rather have as few mobile apps as possible esp if there is a perfectly fine web front end. I do feel that if the overall front end performance is improved the mobile browser experience will follow.
Another thing to report but not sure how to measure is a possible memory leak/resource hogging even after I close the tab. I have a couple of FF threads at the top of htop and my fan running hard even after closing the tab. I had to quit the browser to fully calm the machine down.
I am not able to get the Web Developer Tools to come online enough. I am getting from the FF console numerous Out of memory errors.
I also attached a screenshot from
about:processes
showing that example link above makes a tab process rocket up to 5GB!! This was also a new profile without any addons; so pretty fresh.Putting these here for safe keeping (in case we are wondering about postgres):
https://docs-develop.pleroma.social/backend/configuration/postgresql/
and
https://blog.soykaf.com/post/postgresql-elixir-troubles/
Problem I am seeing seems to still point at client/js, but why not look at DB tuning for scaling as "fe.disroot.org" gets out of beta.
Started to have the same issue just a couple of days ago, though I also noted this slowness to load from the begining of the instance (but I thought it was my internet connection). I had to change the web shortcut URL in order to avoid the landing page which sometimes takes forever to access.
I browsed the upstream dev's instance and did not have the sort of resource utilization as "fe.disroot.org"'s front end nor did the browser hang or hesitate.
@muppeth switching to dev branch fixes performance issues on front end.
https://fe.disroot.org/@muppeth/posts/AEZkNOcy03dwIYyFRw
I would consider this closed, but will let the team officially close it just in case muppeth wants this as a reminder to implement autobuild.
Closing.