38 lines
2.4 KiB
Markdown
38 lines
2.4 KiB
Markdown
# End of support for repl.co & recent issues explained
|
||
|
||
- **期号**: SRE Weekly Issue #413（2024-02-25）
|
||
- **作者**: Tomas Koprusak — UptimeRobot
|
||
- **链接**: https://uptimerobot.com/blog/replit-monitors-recent-issues/
|
||
|
||
## 简介
|
||
|
||
What an interesting cause for an incident: the service your customers have pointed your product at decides to block your requests, effectively DoSing your systems.
|
||
|
||
## 正文
|
||
|
||
As some of you may have noticed, we encountered a snag that led to a brief unexpected pause to your monitors.
|
||
|
||
With the situation now fully under control, we’d like to offer a bit of context and inform you of the next steps regarding repl.co monitors, the culprit of this issue.
|
||
|
||
## What happened
|
||
|
||
As you may remember, we had to [throttle repl.co monitors](https://uptimerobot.com/blog/we-are-throttling-repl-co-monitors/) in the past to prevent the negative impact of repl.co instability resulting in massive & frequent incidents, but keeping the keep-alive option available for free users.
|
||
|
||
Recently, Replit decided to **refrain UptimeRobot users from using our service to ping their repl.co projects** and keep them alive.
|
||
|
||
As a consequence of this decision, Replit removed DNS records for the repl.co domains, which forced our systems into an emergency mode to minimize the effect on other users. There was a 7-digit number of such monitors. This caused some of our monitors to briefly pause.
|
||
|
||
We were able to **resolve the situation quickly**, but to ensure the stability and integrity of our systems moving forward and to comply with repl.co’s request to stop pinging their endpoints, **we’ll no longer be able to support these monitors**.
|
||
|
||
Moving forward, all repl.co monitors will remain stopped and the “Keep-alive mode” feature will be discontinued. Users won’t be able to unpause Replit monitors or create new ones using the repl.co domain.
|
||
|
||
## What’s next
|
||
|
||
For the past few quarters we have been working hard on a brand new monitoring architecture, and we are close to getting it live in beta.
|
||
|
||
This new architecture should be not only more performant and effective, but also much more resilient to even these large-scale DNS, and any other issues.
|
||
|
||
For the immediate term, in response to this situation, we’ve also bolstered our current system to enhance its stability and avoid this type of impact on our service reliability.
|
||
|
||
We apologize for any inconvenience this may have caused and appreciate your support.
|