No issues at all on this front in the West of Ireland. Using 2Gb WiFi at home, 1.2 actual. Using the Discourse app on iPhone/iPad or Safari on the Mac
And several geographically dispersed people have not. There are lots of things this could be that are un-site related (ie Amazon AWS had an outage in the last week – that might be affecting things for some people). Should Adam check? Of course! But it’s starting to feel a bit like piling on at this point.
The “extreme load” messages do worry me, but it’s weird that I’m not seeing them unless I open a lot of topics quickly with Command-clicks. I read every message that’s posted here, and while I may not be the most prolific poster at the moment, I’m certainly near the top. A few other questions, @Simon:
- I’m assuming you’re logged in at all times?
- If you see laggy behavior with reading while logged in, can you try in a private browsing window that’s logged out and see if the same things happen? (You can’t post while logged out, so this will only be useful for trying to identify slow page loads.)
- You’re using multiple networks, which is good, and both are plenty fast. But is it the same Mac that you’re transporting between them? Could you try from another Mac or at least another account? Grasping at straws, I know.
I’ve never seen this error.
Does the site use a content delivery service?
I’d suspect client-side DNS issues.
Of course.
Well yeah, of course, you first want to rule out any client side shenanigans.
It’s on multiple Macs on different networks, different user accounts (macOS, not TidBITS), and irrespective of cookie state or browser.
Two entirely different networks. And it’s a highly intermittent issue, far from persistent.
I’m looking into this more, but I haven’t found any smoking guns. The closest is fairly high memory usage on our DigitalOcean droplet, so I’m talking with my developer if a virtual hardware upgrade might help.
I haven’t seen the “Due to extreme load …” message or noticed any sluggishness since @ace updated the Discourse software.
I am on a terrible network, which I do not control, and am a regular visitor at odd times of day.
I have not had issues.
I wonder about bot activity, since they tend to arrive in swarms, and if possibly a scheduled backup is causing lag for some users.
I’ve talked with my developer about this, and while he made some tweaks to reduce memory usage and free up some disk space, he agreed that upgrading the virtual hardware would probably help. (And it will provide some more breathing room on disk space anyway.) It might involve up to 30 minutes of downtime, and I’ll look into doing it sometime this week.
