Redis 4.x Cookbook

There's more…

If you have checked all the preceding possible reasons for latency issues but still have no clue, you can turn to the help of the internal latency detection.

1. Use the latency watchdog to see what the call stacks looked like when the latency happened:

$ bin/redis-cli config set watchdog-period 500          
OK
$ bin/redis-cli debug sleep 3
$ bin/redis-cli config set watchdog-period 0   
OK
    
8787:signal-handler (1514447053) 
--- WATCHDOG TIMER EXPIRED ---
EIP:
/lib/x86_64-linux-gnu/libc.so.6(+0x14e0eb)[0x7fe381eb40eb]
    
$ tail -n100 log/redis.log   
... 
Backtrace:
bin/redis-server 0.0.0.0:6379(logStackTrace+0x45)[0x46ab25]
bin/redis-server 0.0.0.0:6379(watchdogSignalHandler+0x1b)[0x46abfb]
/lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7fe382141390] ...