Cloudflare 确认 8 月 23 日出现服务中断，并在状态页面上悄然发布了相关信息
2 分•作者: Kal2ef•9 个月前
https://www.cloudflarestatus.com/incidents/pscy7zvw26r1
Cloudflare 的 Discord 机器人昨天 UTC 时间晚上 8:45 发布了这条消息，所以大概是在那个时候发布的：https://discord.com/channels/595317990191398933/895794943182909470/1410726965710688367
我在 8 月 23 日遇到了这些问题，而且我的网站遇到的不仅仅是“延迟升高”。
出现了大量的超时，导致页面根本无法加载，或者部分加载，包括未命中源站的缓存资源。
最近，Cloudflare 一直存在问题，他们会定期将美国的大部分流量路由到更远的数据中心（通常是欧盟）。
他们已经确认这种情况正在发生，因为他们美国的的数据中心无法承受负载，但他们的欧盟数据中心也无法承受。任何使用企业版套餐的用户都会被优先处理，因此这种情况不太可能发生在他们身上。
有时，当这种情况发生时，延迟会再次升高，导致超时，即使是对于缓存资源也是如此。
停机是难免的，对此我表示理解。
但只有当人们在他们的 Discord 上抱怨时，才会被确认——否则，你根本看不到任何信息，除非你去了他们的 Discord，否则你不会知道 Cloudflare 正在努力修复它。
更糟糕的是，他们会在他们的状态页面上悄悄地发布关于影响很多人的停机时间的信息，但他们却专门写了一篇博文来讲述 8 月 21 日发生的停机事件：https://blog.cloudflare.com/cloudflare-incident-on-august-21-2025/
他们在发布什么和不发布什么方面，明显缺乏一致性。
我并不是说这次停机或延迟问题应该专门写一篇博文。
但肯定有比 Discord 更好的沟通方式，除非你只是想隐藏它。
公司经常在他们的状态页面上撒谎，所以对此我并不感到惊讶。但 Cloudflare 试图保持关于停机或其他问题的透明形象，人们经常称赞他们为此所做的详细博文。
这与透明完全相反。
我怀疑他们不在乎，仅仅是因为很多影响都发生在非企业客户身上，但这是一家公司非常糟糕的运营方式。
查看原文
https://www.cloudflarestatus.com/incidents/pscy7zvw26r1<p>Cloudflare's discord bot posted this yesterday at 8:45 PM UTC, so it was published around that time: https://discord.com/channels/595317990191398933/895794943182909470/1410726965710688367<p>I experienced these issues on August 23rd, and my website didn't just have "elevated latency".<p>There were a lot of timeouts where pages did not load at all, or loaded partially, including cached resources that did not hit the origin.<p>Lately Cloudflare has been having issues where they regularly route a significant portion of US traffic over to data centers further away (usually the EU).<p>They've confirmed that this is happening because their US data centers can't take the load, but their EU ones cannot either. Anyone on an Enterprise plan is prioritized, so it's much less likely to happen to them.<p>Sometimes when this happens, there's elevated latency again, which results in timeouts, even for cached resources.<p>Downtime happens, so I can understand that.<p>But that's only been confirmed on their Discord when people complained about it -- otherwise it's nowhere to be seen, and you wouldn't know that Cloudflare is working on fixing it unless you went to their Discord either.<p>To top things off, they silently post about downtime that affects a lot of people on their status page, yet they make a whole blog post about downtime that happened on August 21st: https://blog.cloudflare.com/cloudflare-incident-on-august-21-2025/<p>There's a glaring lack of consistency about what they will or won't post about.<p>I'm not saying that this downtime or latency issues deserves a whole blog post.<p>But surely there's better ways to communicate that than Discord, unless you're just trying to hide it.<p>Companies lie on their status pages all the time, so I'm not surprised about this. But Cloudflare tries to maintain an image of transparency about downtime or other issues, and people often commend them for doing detailed blog posts about it.<p>This is the complete opposite of being transparent.<p>I suspect they don't care just because a lot of the impact is on non-enterprise customers, but that's a pretty horrible way to run a company.