Our systems have been tested time and again. Today they did not perform as designed or expected. All systems are fully operational now but DB contention caused partial API outages earlier today. This is unacceptable. We are making requisite changes.
To all impacted builders and users: we're sorry. We take our operational excellence very seriously and this isn't it. DMs always open if you want to chat.
@henri_stern Hi Henri! Congrats on the load test!!! Is there a reason injected wallets aren’t shown or available as a connection option? IE a “browser wallet” connection option to support the standard browser wallet connection instead of app links to metamask etc
@grntwrght I'm not sure congrats are in order. To answer your q: they are shown but we've found labeling as "browser wallet" is more confusing than actually showing off the injected wallet brand so the user can explicitly pick it. Does that make sense or am i misunderstanding your q?
@henri_stern Still hoping for telegram login to come back up as well
@DuckDegen That’s an unrelated config issue but should be good no?
@henri_stern Hi @henri_stern It seems like a fundamental design and topology decision that high load in one privy wallet provider can cause issues with the rest Is that correct? I don’t understand why it would be necessary for them all to be co-dependent on the same infra like this
@wholisticguy Good Q. In this case, auth issues not about wallets but affected our APIs just the same. We’re making changes accordingly.
@henri_stern Appreciate the transparency. It’s frustrating when trusted systems slip, but acknowledging the issue and committing to fixes is exactly the right move. Followed you 🤝🫡 @henri_stern let’s build together!
@henri_stern The recent issues highlight the importance of robust systems. Continuous improvement is key.
@henri_stern Great takes, but I wouldn't take the brunt on your shoulders. Privy has been a great unlock in this space, but it's still ultimately up to teams to also be responsible for their own resiliency. This isn't the first , nor will it be the last outage. Also saying it's
@henri_stern Thanks for the comms. I know that building systems with this much use doesn’t happen without a hiccup here and there. Keep up the great work 👌
@henri_stern I am proud to see how much @privy_io matters to the space. There are thousands of apps/so-called infrastructures in this space that have never been used as much as privy has been in the last day. breakage due to growth is actually the best problem anybuilders can truly dream
@henri_stern It's all good sir, everyone wanted free money. We still love and use privy
@henri_stern 🫡 if you aren't a privy user, you should be. they build for the users
@henri_stern That's rough Henri we all know stability is king in this space uptime matters big time 📉
@henri_stern proud to work with you and build with you, onwards 🫡
@henri_stern 🫡🫡
@henri_stern Reconocer el fallo y corregirlo rápido es lo que distingue a los equipos que realmente saben de ingeniería. Así se fortalece el sistema.
@henri_stern AWS has world-class SREs and DevOps and a simple config automation update brought down S3 and 50% of the internet in 2017 Five 9's of uptime is the golden standard because there are so many moving parts operationally and technically
@henri_stern embrace the chaos, learn from it
@henri_stern Big brain moment
