If I make more than 10 requests, approximately 3 seconds or fewer apart, the 11th onwards will fail with the following error:
Error:
0: Internal error: Failed to parse IdentityTokenResponse
Location:
crates/bws/src/main.rs:172
Backtrace omitted. Run with RUST_BACKTRACE=1 environment variable to display it.
Run with RUST_BACKTRACE=full to include source snippets.
After 20 seconds or so, it will work again.
At a guess, some sort of poorly handled rate limit? Anyone else seeing this?
Is there any update on this? My primary use case for this is Ansible, and I wrote a simple PoC Ansible lookup plugin for bws, but this bug makes it unusable for anything more than a very minimal number of secrets.
Bump I’d love to start making greater use of bws so I can properly evaluate it and potentially start integrating it into our workflows at work, but this issue is a real blocker for me.
@bw-admin sorry to keep pushing, I know these things take time, but is there any update? Not after an immediate fix, it would just be nice to know where the team is on this one.
Has the team acknowledged the issue?
Is it a deliberate rate limit?
Will the rate limit be increased if deliberate//resolved if not?
I just updated to the latest BWS CLI release (0.3.0) and it seems the error has been updated at least:
Error:
0: Received error message from server: [429 Too Many Requests] {"message":"Slow down! Too many requests. Try again in 1m.","validationErrors":null,"exceptionMessage":null,"exceptionStackTrace":null,"innerExceptionMessage":null,"object":"error"}
Location:
crates/bws/src/main.rs:307
Backtrace omitted. Run with RUST_BACKTRACE=1 environment variable to display it.
Run with RUST_BACKTRACE=full to include source snippets.
That said, this response is clearly server-side, so my update to the latest client is likely immaterial.
As you mentioned that the team is working on a fix for this, I’m assuming that this rate limit was not supposed to be configured so low, and the fix will be to implement an increased rate limit?
Rate limits have been somewhat improved, but they are still both too low, IMO, and entirely undocumented.
For a developer-focused product, I find this to be pretty lacklustre at best. I really hope Bitwarden reconsiders this as I really like BWS, but this is a very frustrating problem, and I’m sure many people will be less patient than I when faced with such a problem.
I’ve been using bws-cache for Ansible lookups. It’s quite unfortunate that this is necessary, and it may well not fit all use cases, but it has helped to alleviate the issue.