You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Apr 27, 2019. It is now read-only.
We have multiple servers with our application (which using playfab nuget) + with consul.io agent on the same server and additional cluster of consul.io servers for high availability
Anyways, once in a few days we getting exception on Non leader application:
Consul.ConsulRequestException: Unexpected response, status code BadRequest: Invalid wait time
Also,
I see in consul agent logs:
2019/03/12 04:54:08 [DEBUG] http: Request GET /v1/kv/service/scm/leader?index=934475&wait=1.999s (1.9967478s) from=127.0.0.1:56777
2019/03/12 04:54:08 [DEBUG] http: Request GET /v1/kv/service/scm/leader?index=934475&wait=ms (0s) from=127.0.0.1:56777
Usually we try to acquire lock every 2 seconds, so we see every to seconds GET log with wait=...s..
But when problem happens we see 2 GETs when second has wait=ms, which causes that exception.
I checked playfab code, public async Task Acquire(CancellationToken ct) method. but I didn't find good explanation why it happens? Some rare race condition?
10x in advance
The text was updated successfully, but these errors were encountered:
Actually the best fix is in code:
class Lock
public async Task Acquire(CancellationToken ct)
Instead of if (elapsed > qOpts.WaitTime)
it should be if (elapsed >= qOpts.WaitTime)
actually it looks like the only place where qOpts.WaitTime is updated and could be 0.
So i would like to do pull request with that fix
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Hi,
We are using this package (version 0.7.2.6) for Leader election using distributed lock.
It is based on blog : https://codeblab.herokuapp.com/high-availability-and-leader-election-with-consul-c-part-1/
It works pretty fine (thanks for great work).
We have multiple servers with our application (which using playfab nuget) + with consul.io agent on the same server and additional cluster of consul.io servers for high availability
Anyways, once in a few days we getting exception on Non leader application:
Consul.ConsulRequestException: Unexpected response, status code BadRequest: Invalid wait time
Also,
I see in consul agent logs:
2019/03/12 04:54:08 [DEBUG] http: Request GET /v1/kv/service/scm/leader?index=934475&wait=1.999s (1.9967478s) from=127.0.0.1:56777
2019/03/12 04:54:08 [DEBUG] http: Request GET /v1/kv/service/scm/leader?index=934475&wait=ms (0s) from=127.0.0.1:56777
Usually we try to acquire lock every 2 seconds, so we see every to seconds GET log with wait=...s..
But when problem happens we see 2 GETs when second has wait=ms, which causes that exception.
I checked playfab code, public async Task Acquire(CancellationToken ct) method. but I didn't find good explanation why it happens? Some rare race condition?
10x in advance
The text was updated successfully, but these errors were encountered: