Remix.run Logo
MoroL 7 hours ago

Sorry everyone. My hosting was not prepared for the amount of traffic while I migrate you can find the explorations on Twitter as well.

https://x.com/lukas_moro/status/1829487148078412019 https://x.com/lukas_moro/status/1838207092471050645 https://x.com/lukas_moro/status/1847299759603699906

leerob 7 hours ago | parent [-]

(I work at Vercel) It seems like you possibly have a spend cap on, which would automatically pause your site when a certain amount of spend is hit. Vercel is working correctly in this case.

If you are hosting a lot of video files, I would recommend using Vercel Blob (object storage). This is a better fit for larger volume assets like images or videos, versus "fast data transfer" as you mentioned in another comment for critical assets like stylesheets or scripts.

Happy to help out if you have questions, email is lee at vercel dot com.

xyzzy_plugh 6 hours ago | parent | next [-]

This is fine content for private communications with your customer but rubs me the wrong way posted here. It doesn't read to me that your customer is being negative about Vercel in any way, so to me it feels arrogant to come here and say "well actually you can't afford us."

If you feel differently, and think it's acceptable to do so, why not instead say that on the error page? "Customer has exceeded their budget. Please add additional funds."

If you don't think such messaging is appropriate, then I'm curious why you think doing it here would be?

If you want to be truly helpful to your customer, you would consider raising or temporarily uncapping their traffic as a gesture of good will.

leerob 5 hours ago | parent [-]

The “deployment paused” screen is shown publicly when hard spend limits are reached for a project. These limits are configured by the user.

Spend controls aren’t necessarily about affordability, it’s often for peace of mind (similar to a fixed price server, or a disposal card with a limit).

I also don’t view them as being negative on Vercel (we likely could have alerted them better, as it seems this caught them off guard).

Their traffic isn’t capped and they can change this if they prefer. But I’m guessing there’s some large asset causing unexpected usage, which is why I offered my email. Happy to walk through it with them and figure out a path to optimize.

voytec 6 hours ago | parent | prev [-]

Is ERR_SSL_VERSION_OR_CIPHER_MISMATCH the effect of this cap? If so - that's a really weird approach at blocking a website.

    > openssl version
    OpenSSL 3.4.0 22 Oct 2024 (Library: OpenSSL 3.4.0 22 Oct 2024)
    > openssl s_client -connect www.lukasmoro.com:443
    Connecting to 2a06:98c1:3120::b
    CONNECTED(00000005)
    00E167F201000000:error:0A000410:SSL routines:ssl3_read_bytes:ssl/tls alert handshake failure:ssl/record/rec_layer_s3.c:908:SSL alert number 40
leerob 5 hours ago | parent [-]

No, I am referencing the “deployment paused” page.