image/svg+xml Follow

I have a & instance running on a that i can access from the internet. It's slow. How do I figure out what is the bottleneck?

@mprv
The biggest culprit on raspberry pi's is the disk io speed. This is due to the fact that the ethernet card and usb ports share bandwidth so if you're writing to a usb drive while accepting traffic over ethernet, performance is limited and the built in micro sd card slot has very slow data transfer rates.

Here's a page with some tips for how to improve it github.com/nextcloud/nextcloud

@ksteimel
Thank you, I've done all I can from that list.
I am more interested in a more specific test: how do I understand if the problem is my internet connection (which would not be fixed by using more powerful hardware) or the Raspi? Do you have any idea?

@mprv
Is it running on your home network with port forwarding set up at your router?

@ksteimel yes. I used no-ip to get a free domain name.

@mprv
If you access your nextcloud instance using the local ip from inside your home network and then compare that to using an outside network that will tell you whether the bottleneck is the network or the rpi.

Typically what I do is I see how long the page takes to load when my phone is on the local wifi, then I turn off wifi and see if it's perceptually longer over cellular. You want to use private browsing though so that you don't use your browser's cache.

@ksteimel
I didn't think about the cache issue, that's a good point.

@mprv @ksteimel Have you tried dstat or iostat to monitor disk I/O? sar is also a good thing to see historical data on system utilisation.

iptraf can show you bandwidth utilisation.

Also check top -- is the load average high, is something eating up all the memory?

Those are just a few thoughts. I hope that's helpful! :)

@michelamarie @ksteimel
Thank you for the suggestions! I checked top and every time I interact with the web interface the load shoots up. I'm not sure how to interpret those data, though.

The results of iostat show that the microSD I'm using is more than adequate.

I haven't tried with the other tools, but I suspect that the bottleneck is combination of cpu, memory and bandwidth (in this order).

I'm considering buying an intel NUC, since I want to self host as much as I can. Thank you again!

Sign in to participate in the conversation
Scholar Social

Federated microblogging for academics

Scholar Social is a microblogging platform for researchers, grad students, librarians, archivists, undergrads, academically inclined high schoolers, educators of all levels, journal editors, research assistants, professors, administrators—anyone involved in academia who is willing to engage with others respectfully.

We strive to be a safe space for queer people and other minorities in academia, recognizing that there can only be academic freedom where the existence and validity of interlocutors' identities is taken as axiomatic.

"An academic microblog that you can be proud to put on the last slide of a presentation at a conference"

"Official" monthly journal club!

(Participation is, of course, optional)

Scholar Social features a monthly "official" journal club, in which we try to read and comment on a paper of interest.

Any user of Scholar Social can suggest an article by sending the DOI by direct message to @socrates@scholar.social and one will be chosen by random lottery on the last day of the month. We ask that you only submit articles that are from *outside* your own field of study to try to ensure that the papers we read are accessible and interesting to non-experts.

Read more ...