bottleneck
bottleneck copied to clipboard
ReplyError: ERR SETTINGS_KEY_NOT_FOUND when use in Clustering mode with Redis 7.x
Hi!
First, thank you for this great library 🙂
I tried running this library in Cluster mode with Redis 7.x, and I got the following error:
ReplyError: ERR SETTINGS_KEY_NOT_FOUND
command: 'EVALSHA',
args: [
'f00a66a4159e8488fd6e3b8ccba92a8c602f2062',
8,
'b_gladys-gateway_settings',
'b_gladys-gateway_job_weights',
'b_gladys-gateway_job_expirations',
'b_gladys-gateway_job_clients',
'b_gladys-gateway_client_running',
'b_gladys-gateway_client_num_queued',
'b_gladys-gateway_client_last_registered',
'b_gladys-gateway_client_last_seen',
1662654440828,
'4gydkiwu33',
'0',
'1'
],
code: 'ERR'
}
If I use an older version of Redis, it works fine.
Any idea of what changed with this version of Redis ?
Just ran into this as well
same issue here as well
Getting this same error as well:
ReplyError: ERR SETTINGS_KEY_NOT_FOUND
at parseError (/usr/api/node_modules/redis-parser/lib/parser.js:179:12)
at parseType (/usr/api/node_modules/redis-parser/lib/parser.js:302:14) {
command: 'EVALSHA',
args: [
'8bd3e064b718321194a8df49c69e48fd0ffb4dcd',
8,
'b_builtwith_throttler_settings',
'b_builtwith_throttler_job_weights',
'b_builtwith_throttler_job_expirations',
'b_builtwith_throttler_job_clients',
'b_builtwith_throttler_client_running',
'b_builtwith_throttler_client_num_queued',
'b_builtwith_throttler_client_last_registered',
'b_builtwith_throttler_client_last_seen',
1665586647495,
'b9zz33fpice'
],
code: 'ERR'
}
We're also experiencing it with bitnami/redis:7.0.4-debian-11-r4
, any workaround? @SGrondin
Same here:
ReplyError: ERR SETTINGS_KEY_NOT_FOUND
command: 'EVALSHA',
args: [
'8bd3e064b718321194a8df49c69e48fd0ffb4dcd',
8,
'b_awql_throttler_settings',
'b_awql_throttler_job_weights',
'b_awql_throttler_job_expirations',
'b_awql_throttler_job_clients',
'b_awql_throttler_client_running',
'b_awql_throttler_client_num_queued',
'b_awql_throttler_client_last_registered',
'b_awql_throttler_client_last_seen',
1665734323793,
'0efrswlefrmj'
],
code: 'ERR'
}
We're experiencing this on an elasticache redis instance running 6.2.6, any updates on this? @SGrondin
It's probably happening because of the heartbeat check which validates that keys exist on redis. But since Groups timeout kills the limiter after timeout those keys don't exist anymore. https://github.com/SGrondin/bottleneck/blob/b83528333ba4d27cf70b81cc2be12e09d7ff692f/src/Scripts.coffee#L84
The error usually goes away once the group_check.lua
runs so a quick hack would be ignore the error for a while until it goes away
Also ran into this. Between this issue and ETIMEDOUT errors, it's been very difficult to get ioredis cleanly working w/ ElastiCache 😢