-
Currently, the `tower-reconnect` middleware tries to reconnect immediately if an error occurs _while_ connected, but propagates all errors if an error occurs while _connecting_. We should modify this …
hawkw updated
4 years ago
-
Now that we have signed peer records in libp2p, we should use them for peer addresses in the DHT.
**Need to take these to completion:**
https://github.com/libp2p/go-libp2p/pull/747.
https://git…
-
I changed the pad_symbol as left_pad_symbol, right_pad_symbol and add start_pad_symbol in KneserNeyLM, but there still another eroor. We may use log function with a negative value,but why it was neg…
-
Right now, the current behavior for reconnection to sockets will cause an infinite loop of retrying connections if the server goes down. To mitigate issues like this, our providers should support some…
-
## What is the current behavior?
The retry behavior when snowflake responds with a 503 is not controllable. It currently retries 10 times with a jittered 15s backoff.
## What is the desired behav…
-
We're moving a bit too fast in some cases so we should implement some backoffs
https://github.com/App-vNext/Polly should be a good solution to backoff
-
Similar to other metrics we have during upload, we should include at least:
- number of successful uploads
- number of failed uploads
- upload backoffs
Unlike some of the other upload `remote` m…
-
Relevant: https://github.com/redis/redis-py/blob/master/redis/retry.py
Backoffs: https://github.com/redis/redis-py/blob/master/redis/backoff.py
Examples: https://github.com/redis/redis-py/blob/d…
-
-
/area quality
/kind enhancement
/priority 3
**What would you like to be added**:
MCM needs to have a way to configure API requests frequency.
**Why is this needed**:
On failure to create an…
bg502 updated
2 months ago