It always hangs on docker_manager with the following error
Failed to load resource: the server responded with a status of 429 (OK)
docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:11 Error: OK
at o (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:11)
at Object.e.default.trigger (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:16)
at h._onerror (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:16)
at d (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:16)
at docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:11
at r.invoke (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:4)
at Object.r.flush (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:4)
at Object.n.flush (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:4)
at Object.a.end (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:4)
at Object.a.run (docker-manager-vendor-e62406a2f5676e280b40622751069e76.js:4)
Could be standard rate limiting stuff. In the network tab, you should be able to see more details about that 429 since we send the description of the exception back to the client.
So with it possibly being a ratelimit, should that be excluded for Admins, especially on the Upgrade page? Or provide a way to have it redo the check for a given package/component?
As the way it sits, I can’t do an upgrade via the web anymore (which isn’t critical to me as it is a ‘sandbox’ instance).
FYI, I updated the requests per second and that solved my issue of using the web interface for upgrades. Since it is a sandbox, I don’t mind doing that, realizing that may not be recommended for typical production installs.
Hm, I don’t understand how this is the solution other than it identifies the cause of the problem. Could you explain how I can have more than 11 plugins installed?
(Strangely, the rate limit doesn’t always seem to trigger. Perhaps because update requests are not re-sent within a certain time period so that when I try again after a short while, the remaining requests can get sent without triggering the rate limit?)
The correct solution would be to have the page make fewer requests - if that’s impossible (I forget how the plugin has been built) then it should be throttled so that only a few plugins are checked simultaneously.
So you’re saying the docker-manager plugin needs to be fixed and there’s nothing I can do as a user? I’m sure I’m not the only one with more than 11 plugins installed and even for me it has been working most of the time. I don’t quite understand why I sometimes get the 429 error and sometimes I don’t. And when I get it, it sometimes works when I just re-try immediately. Sometimes it works after re-trying after a while and sometimes I just can’t get it to work. (And by “work” I mean get all plugins checked so that the “Update to latest version” button turns blue and becomes clickable.)
The reason you’re seeing it happen sometimes and not others has to do with the speed at which the server and your internet connection can respond, which is variable. If it happens too fast, you will be throttled.
I’ve updated docker_manager to now only check for plugins sequentially. It’s slightly slower but it will prevent this bug which I consider quite severe: