问题描述
昨天Chrome发布了最新版本(64.0.3282)之后,我现在收到了这个错误:
我在Docker容器中运行大约2000次硒测试,我发现这个失败的速度大约为1为100.根据我的说法,没有可重复的步骤 - 每次迭代失败的测试都不相同。我更新到最新的Chromedriver(2.35),但这似乎没有任何影响。我以前使用Selenium 2.41,但已经更新到最新版本(3.8.1),希望它可以帮助...它没有。我完全不知道为什么会发生这种情况。有没有其他人注意到这一点?是否可能是Chrome最新版本的bug?
预先感谢您提供的任何帮助。
看起来最新的Chrome版本存在问题。如果未设置disable-gpu Chromeoption集,则渲染器有时会超时。 Google解决这个问题的解决方法是,在ChromeOptions中添加--disable-gpu属性。
编辑:这样可以减少频率的事件发生,但仍然在发生。After Chrome released their newest version yesterday (64.0.3282), I am now receiving this error rather sporadically:
I'm running about 2,000 selenium tests within a docker container and I see this failure at a rate of about 1 in 100. There are no reproducible steps as far as I can tell- the tests that fail are different with each iteration. I updated to the newest Chromedriver (2.35), but that didn't seem to have any effect. I was previously using Selenium 2.41, but have updated to the newest version (3.8.1) hoping that it might help... it did not. I'm completely at a loss as to why this might be occurring. Has anyone else noticed this? Is it possibly a bug with Chrome's newest release?
Thank you in advance for any help you may be able to provide.
It looks like there was an issue with the newest Chrome release. Without the disable-gpu Chromeoption set, the renderer will occasionally timeout. The workaround until Google fixes this (if they do fix it at all) is to add the --disable-gpu attribute to the ChromeOptions.
EDIT: This reduced the frequency of occurrences, but it is still happening.
这篇关于Selenium超时接收来自渲染器的消息的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!