Dear Developers, Artists, Community,
Thank you heartfelt for the marvel, art…
There’s the following possible case regarding image hosting.
The following URL was set to a post once posted:
- https://lemmy.world/pictrs/image/410efc36-1797-4d02-b771-ec608c21a64e.png?format=webp
The related post:
- https://lemmy.world/post/49812750 (Nancy Kovack (1963)…)
Once the post was loaded shortly after its publishing, in a local client, the client returned a “no image” placeholder.
When checked the raw response, the endpoint sent the following:
{
"error":"unknown",
"message":"Request error: error sending request for url (http://pictrs:8080/image/process.webp?src=410efc36-1797-4d02-b771-ec608c21a64e.png&thumbnail=256): operation timed out"
}
It seems like a container/virtual machine, with the custom hostname of pictrs is not available at certain times on port 8080, as if restarting shortly or too busy right the moment the post is created, possibly proxied within containers/local network where custom DNS resolved of Docker/Podman/etc. or similar systems exist. In result, “times out” on DNS or another transfer protocol context.
With some time passed, a few minutes, with the post reloaded for another request, the image was successfully received!
Please stay safe, and I wish you success, stability, and peace…
Best and kind regards
Roger that! Thank you very much for the response, care, and time, dear MrKaplan!
I am sorry for the long timed response…
If the changes are upstream, I haven’t yet had a chance to verify it, and will report once found. It has been happening randomly, and the exact condition was not defined.
@[email protected] , just to clarify, did you encounter the case since the change?
Meanwhile, sincere gratitude for the ineffably magnificent art you do!
every now and then, usually when i’m preparing the post backlog but not really frequently.