Hi, this error is becoming really annoying. As i understand, during that time there is some updating code or something. But this actually mean that at short time, blockchain is down. If string of processing happening at that time and it cuts when halfway done, than it can result in loss of data or funds. And to prevent that, than every step has to be coded so that process can be recovered wherever it stopped, adding those steps will make app slower.
Will those outages haunt us forever or this is just temporary, last 30 days i have seen many of those.
1 Like
I’m having more of these, not only for the initial loading of the page, but also for assets requested after load, sometimes with (e.g.):
’ Uncaught (in promise) ChunkLoadError: Loading chunk 395 failed.
(error: https://jtjkb-miaaa-aaaai-qpffq-cai.icp0.io/395.index.js)’
I also get either 500 or 503 errors reliably when i try to upload files above a certain size (~20 MB) but will be making a separate thread for this
I’m also beginning to think it’s related to my system config, because others accessing the canister do not have the same problem. Wish I knew more…
I have no idea, PersonalDao gives 503 error, Internet Identity login on ICPEx was so slow, swaps dont work. Whatever Dfinity is updating is kicking apps offline and making them to not work properly.
Hello @Henn91 and @willguest
Sorry for the inconveniences! We had one HTTP gateway with a higher error rate starting from 19:00 UTC. We removed it from the loadbalancer and looked into.
What would help us a lot to investigate is to provide us with the IP address of the remote server and the request ID. You can find both when opening the networking tab in the developer tools of your browser.
We have alerts that trigger if the error rate of single HTTP gateways or overall pass some limit. However, the overall error rate was still low: normally, we see about 1 503 per second, starting from 19:00, it increased to 3 per second. This is around 0.05% of all responses.