Replies: 6 comments 13 replies
-
Hey, you will need to submit more tangible evidence and analysis - what specific step is taking more time? The best way is to analyze the debug log files and see what's taking longer. The evidence you've provided in #14 is not detailed enough to work with, unfortunately. I have also run some performance metrics on our infrastructure and didn't detect performance issues. |
Beta Was this translation helpful? Give feedback.
-
[email protected] + [email protected]
[email protected] + [email protected]
|
Beta Was this translation helpful? Give feedback.
-
Will it have something to do with the docker environment? |
Beta Was this translation helpful? Give feedback.
-
At present, it should be the problem of slow response of proxy service |
Beta Was this translation helpful? Give feedback.
-
Hey, @jiadesen and @mirobo thanks a lot for sharing all the information - I can imagine that it's difficult to debug and troubleshoot the performance issues. Let's try to focus on specific issues you're facing and not mix evth together. If you suspect that cy2 proxy implementation is slow, please try to collect specific data related to the proxy performance. I need samples of specific evidence of the slowness in cy2 - i.e.:
Also, when relying on debug output, please get familiar with how the timing of So far you've been posting the debug logs with no interpretation and no real evidence. It can be your network setup or internal cypress issues that are not related to cy2 and I am finding myself asking the same questions again and again. There are about 40K weekly downloads of [email protected] - if there's a real slowness I'd imagine much more complaints. To summarize - please provide samples of specific evidence of cy2 slowness - i.e.:
As a side note, we don't really have any different way to integrate with cypress at the moment. |
Beta Was this translation helpful? Give feedback.
-
In Gitlab CI I do:
Log cy2log-clean.txt From the above log I searched for long running tasks with "+[0-9]+m[^s]" and copied the starting point of the measurement, the stuff that happens in between and the ending point of the measurement (same logger context). Total job duration in GitLab is 15 minutes and there are 6 pairs of "cypress:server:cloud:api request" + "cypress:server:cloud:api response" and between each request and response it takes 2 minutes which should account for 12 minutes of the total duration of the job. Before the Cypress/cy2 upgrade the job took around 2-2:30 minutes (Cypress 10.7.0). Within each of the pairs there is "network:agent addRequest called" + "network:agent got family" which also takes 2 minutes. So I guess the delay comes from there. POST /runs
POST /instances
POST /instances/.../tests
POST /instances/../results
POST /instances/.../stdout
POST /runs/.../instances
|
Beta Was this translation helpful? Give feedback.
-
The more test files, the more obvious the effect. eg:
[email protected] + [email protected]
[email protected] + [email protected]
#47
Beta Was this translation helpful? Give feedback.
All reactions