My Nextcloud has always been sluggish — navigating and interacting isn’t snappy/responsive, changing between apps is very slow, loading tasks is horrible, etc. I’m curious what the experience is like for other people. I’d also be curious to know how you have your Nextcloud set up (install method, server hardware, any other relevent special configs, etc.). Mine is essentially just a default install of Nextcloud Snap.
Edit (2024-03-03T09:00Z): I should clarify that I am specifically talking about the web interface and not general file sync capabilites. Specifically, I notice the sluggishness the most when interacting with the calendar, and tasks.
mine was really sluggish for a long time, then I saw someone in here explaining their similar issue and their fix. I don’t have the post link, but it was related to DNS settings. Basically for some reason using my pihole dns made only nextcloud sluggish, the fix suggestion was to use 1.1.1.1, which worked. Now, it is a pretty fast nextcloud.
So on your Nextcloud server you use an external DNS and it greatly sped up you nextcloud? Because I noticed a few years back mine got slow and I cannot figure out why. It was about the time I enforced pihole dns with pfsense. I might need to try this.
That would make sense if the cause is some looping from hanging DNS lookups. Someone should (and likely has) notified the devs about this.
Another possible solution, from https://help.nextcloud.com/t/server-hangs-and-then-is-fine-for-a-bit-then-hangs-again/153917/16
I’m going to have to give this a shot tonight, need to make a pfsense rule to allow the server to get out and then change its DNS. Regarding php, my current config is the following because I have over 64gigs of ram and went through great length to get Nextcloud to cache MORE into ram:
pm.max_requests = 50000 #set higher, the process is recyled after 50k calls to prevent memory leaks pm.max_children = 1000 pm.start_servers = 60 pm.min_spare_servers = 30 pm.max_spare_servers = 120
I am very happy with mine and have only ever had one hiccup during updating that was due to my Dockerfile removing one dependency to many. I’ve run it bare metal (apache, mariadb) as well as containerized (derived custom image, traefik, mariadb). Both were okay in speed after applying all steps from the documentation.
Having the database on your fastest drive is definitely very important. Whenever I look at htop while making big copies or moves, it’s always mariadb that’s shuffling stuff around.
In my opinion there are 2 things that make nextcloud (appear) slow:
-
Managing the ton of metadata in the db that is used by nextcloud to provide the enhanced functionality
-
It is/was a webpage rendered mostly on the server.
The first issue is hard to tackle, because it is intrinsic and also has different optimums for different deployment scales. Optimizing databases is beyond my skillset and therefore I stick to the recommendations.
The second issue is slowly being worked around, because many applications on nextcloud now resemble SPAs, that are highly interactive and are rendered by your browser. That reduces page reloads and makes it feel more smooth.
All that said, I barely use the webinterface, because I rarely use the collaboration features. If I have to create a share I usually do that on the app because that’s where I send the link to people. Most of my usecase is just syncing files, calendars and contacts.
Containers run on “bare metal”…
-
I run Nextcloud on an old laptop (i5-8500h) and tbh I find it super fast and responsive. I’ve barely done any tinkering or customization
Pretty sluggish for me as well. Bare metal install with Apache, PHP 8.3, since a few days PostgreSQL and the whole Redis memcached opcache whatevercache stuff. Next step would be to check if the AIO Docker is the magical thing that makes it fly.
Some 8 core CPU I’m too lazy to look up, 16 GB RAM and two HDDs. SSDs would probably help, I guess.
Docker, using the
nextcloud:stableimage (not-all in-one) with postgres, behind nginx, and finally ZFS with 2x modern HDDs for storage. I run the stock apps plus a small handful, and have carried the same database through many versions over the last 5 years.It’s usable, but definitely not snappy.
The web interface for files is fine. Not instantaneous at all but not a huge problem. I have about 1TB of files (images and videos) in one folder, then varying files everywhere else. I suspect that the number of files (but probably not the size) is causing the slowdown.
Switching to, for example, the notes app is incredibly slow, and the NC Android app is just as bad.
Same.
I’ve always run Nextcloud as a docker behind an NGINX/Let’s Encrypt proxy and login sometimes takes over a minute, even if I access the Nextcloud docker directly without the proxy. It’s a very frustrating experience to use a self hosted Nextcloud.
Which docker did you use, out of curiosity?
This one: https://hub.docker.com/_/nextcloud/
I stopped using it because it has an extremely complex protocol, with very large bloat that increases with the number of files, and incredibly sensitive to latency.
When I stopped syncing directories because they would take days to upload and started compressing them so they would finish in 10 minutes, I decided it had to go. (Oh, and it’s extremely sensitive to network problems too.)
I still use Nextcloud for syncing documents and other basic stuff that is relatively simple. But I started getting glacial sync times consuming large amounts of CPU and running into lots of conflicts as more and more got added. For higher performance, more demanding sync tasks involving huge numbers of files, large file sizes, and rapid changes, I’ve started using Syncthing and am much, much happier with it. Nextcloud sync seems to be sort of a jack of all trades, master of none, kind of thing. Whereas Syncthing is a one trick pony that does that trick very, very well.
@Kalcifer @selfhosted it’s quite slow for me. AIO docker setup
It was way worse when I tried the snap tho
As long as it’s faster than the snap, it’s worth it to me 😜
Mine runs very smooth. I guess it’s a matter of hardware or Ressource allocation?
I have a fujitsu thin client with a 4 core Celeron CPU and 8gb RAM. Nextcloud and Maria DB run in separate containers in proxmox.
I am the only user. I have only office, face recognition and maps installed. Other than that I use calender, contact and Foto sync.
Nextcloud pleases A LOT 10% of it’s users. Those 10% are composed by tech savvy people, coders and developpers that spent countless hours tinkering with their instance.
I’m one of the 90% left. Despite really wanting to use nextcloud and trying to set it up correctly for 2 years, I finally gave up and I feel much happier in my life, in my work, with my family and friends, and they thank me for that.
Now I just recommend Owncloud or seafile. They’re both really easy to install and just work out of the box.
Out of habit and convenience, I keep a nextcloud running on oracle free tier just for what it’s good at: caldav and contacts.
Never had an issue with mine. And running fine. Only thing I have done is use mariadb.
I just moved my files from nextcloud to seafile, founded that I don’t really need chat, calendar, tasks and other things, only need to store files and have it synced between my devices.
Nextcloud works for my small company and I’m not going to change it for now.
The out of the box experience of the containerized nextcloud is actually really bad. Had it running bare metal with apache and it was way faster.
But have you tried the official AIO docker compose file? Basically copy the redis stuff from there and you are good to go.
Containers run on “bare metal”.
Not in this context. Bare metal means all packages and services installed and running directly on the host, not through docker/lxc/vms
Yes - in this context containers run on bare metal. They run directly on the host. They even show up in the host’s process list with PIDs. There is no virtual machine between an executable running in a docker image and the CPU on the host.
Have you read my comment? It’s about where the packages and services are installed.
In this case, they’re installed in the container, not on the host
What is it you think the “metal” is in in the phrase “running on bare metal?”
Your comment is irrelevant. Who cares in what directory or disk image the packages are installed? If I run in a “chroot jail” am I not “running on bare metal?” What if I include a library in /opt/application/lib? Does it matter if the binaries are on an NFS share? This is all irrelevant.
The phrase means to be not running in any emulation. To answer my question above - the “metal” is the CPU (edit: and other hardware).
edit2: I mean - it’s the defining characteristic of containers that they execute on bare metal unlike VMs and (arguably - I won’t get into it) hypervisors. There is no hardware abstraction at all. They just run natively.
It’s just what it means in this specific context.
They’re not running directly on the host, with directly meaning directly.
If you go by definition, I agree with you, but the definition is not always the thing to go off of.
It’s all about where the packages and services are installed
No. Your packages and services could be on a network share on the other side of the world, but where they are run is what matters here. Processes are always loaded into, and run from main memory.
“Running on bare metal” refers to whether the CPU the process is being run on is emulated/virtualized (ex. via Intel VT-x) or not.
A VM uses virtualization to run an OS, and the processes are running within that OS, thus neither is running on bare metal. But the purpose of containers is to run them wherever your host OS is running. So if your host is on bare metal, then the container is too. You are not emulating or virtualizing any hardware.
Here’s an article explaining the difference in more detail if needed.
More specifically, the container is run on bare metal if the host is running on bare metal. You are correct in this thread, not sure why you’re being downvoted. I guess people don’t know what virtualization technology is or when it is used.
If the nextcloud container is slow, it’s for reasons other than virtualization.
not sure why you’re being downvoted.
LaNgUaGe EvOlVeS. 🙄
Wait what? I’m saying what you said is correct. Am I the one who’s confused here?
Edit: oh maybe you meant that’s the excuse people give for being wrong? lol
Yes, that’s the excuse I’ve been getting. Sorry to be vague.
Now I just recommend Owncloud or seafile. They’re both really easy to install and just work out of the box.
Which one is lighter on your opinion?
Lighter, I dont know. Faster, I’d say owncloud. YMMV
What hardware are you running it on? I set mine up originally on a raspberry pi 3b and the web interface was very slow, but upgrading it to a RPi4 with 4GB RAM made a massive difference. Though I suspect some of that was that the data and database were being stored on an external SSD in both cases, the RPi4 had a usb 3 interface and dedicated Ethernet, but the 3b had a single USB 2 bus to share between the Ethernet and SSD.
MIne is on a RPi4 4 GB as well (the AIO container) with two SSD and the performance of the frontend are meh… But I seldom use the web part anyway and mainly rely on the desktop client/android App, which work just fine.
Mine is nice and quick in regards to the web interface and general functions. However I run it on a server at home and my upload speed isn’t the best, so if I need to pull a larger file (Files On Demand enabled) then obviously the transfer speed of the file is a bit sluggish.
Hosted on a VM with 16GB RAM, 4 cores. Using the NextcloudAIO docker deployment option, all behind an Apache reverse proxy (I have a bunch of other services on another VM that all have reverse proxy access in place as well).
Did you do anything special configuration-wise, or did you, more or less, just deploy the AIO docker as-is?
Nothing too special, just had to do some fiddling to get the Apache reverse proxy working correctly. Now I believe they have a pre-made example for it, but back then they only had nginx. I stick with Apache because that’s still what I know. Might start learning nginx, but my main work isn’t in web stuff.
Quite fast.
KVM/libvirt VM with 4GB RAM and 4vCores shared with a dozen other services, storage is not the fastest (qcow2-backed disks on a ext4 partition inside a LUKS volume on a 5400RPM hard drive… I might move it so a SSD sometime soon) so features highly dependent on disk I/O (thumbnailing) are sometimes sluggish. There is an occasional slowdown, I suppose caused by APCu caches periodically being dropped, but once a page is loaded and the cache is warmed up, it becomes fast again.
Standard apache + php-fpm + postgresql setup as described in the Nextcloud official documentation, automated through this ansible role
I run linuxserver.io docker container, disabled almost all apps and its been running rock solid and quite fast on old celeron. It takes 3-5 sec to open a web page, but I mostly use desktop/android app anyway
Configuring a Redis cache really helps in my experience.
But I also recently noticed something odd: it works quite well on my usual internet connection, but when I traveled abroad it became excruciatingly slow, more so than the admittably worse mobile connection would have let me assume. Something about it seems to require a relatively stable internet connection on the client side it seems.
That might be due to your ISP’s routing and interconnects. They usually have good routes to big services and might lack good connections between home users in different countries or on different continents.
