-
-
Notifications
You must be signed in to change notification settings - Fork 215
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Should the snap offer the option to use APCu for local cache? #2663
Comments
@pachulo we have a large user base running "limited memory" devices (SBC's) so this should definitely be optional.
According to the recommendation, it could increase performance on "enhanced memory" machines. That would be desirable. |
Yeah, it makes sense that we'd be using redis given our start on the Nextcloud box. However, while I agree with @scubamuc that continuing to support such users is important, I just checked the metrics this morning: literally 95% of our users are amd64. If this would enhance performance for those users, it might be worth investigating as long as we can figure out a way for it to not hurt the other 5%. |
OK, I found more information:
Could we configure this based on available RAM?🤔 P.D. |
No, only Apache is supported:
That's the reason we went this direction in the first place. NGINX is definitely lighter-weight. That's one of the reasons we've been so careful about not opening up Apache configs: we can switch to NGINX in the future with no one being the wiser if we feel it necessary/worthwhile. |
Only 5% running "limited memory" devices (SBC's), interesting...
I would be remiss if I hadn't mentioned these users? |
I'm one of such users, so I can't agree more! 😄
What do you think would make more sense @kyrofa :
|
Um I am a bit confused I am totally unfamiliar with those stuff so is there a way to tune the server performance because I have 256gb ram 24 core but I had closely a crash on my device that was accessing the Nextcloud instance that is on the high performance server it were looking for me that the access device loads the hole stuff not the server where Nextcloud were running on especially when I uploaded 361 images at once my device from where I was uploading were stuck for 20sec! That was mind blowing for me ! :/ |
Any advice would be great! :) |
@alfi4000, you have a powerful machine by the sounds of it... this PR is a development discussion and is not the right PR for your issue... I guess. @pachulo is considering some default memory cache changes for the Snap. as far as your performance issues are concerned, its not as simple as tuning here or there... you'd have to study your system logs and performance monitors carefully to pinpoint the bottleneck. That's admin stuff and not necessarily a Nextcloud snap issue. there may be some knobs to turn if we knew more about your system. but we're trying to make Nextcloud snap work for everyone "out of the box". So we're more concerned with stability than performance... especially since we don't know what device is being used. basically advice for system tuning your machine would be going way beyond what we're trying to do here... also this PR would be wrong for questions about your device. So I'd suggest you make a new PR with your performance issue. |
Ok |
After reading the Recommendations based on type of deployment on the Memory caching page of the nextcloud manual, I was wondering:
memcache.local
instead of Redis?I haven't found any real benchmarks or anything, just that it should be faster but consume more RAM, but as it is also the official recommendation, I just wanted to understand better why we use Redis instead.
The text was updated successfully, but these errors were encountered: