Replica: constantly behind master, some spare keys

Hello,

I am using redis replication scheme for a long time, but the same thing with keydb seems to be very weird. Documentation states that replica is initially set up only with replicaof clause, exactly as in redis, but the replica set up like this just seems to be malfunctional:

Master:

Keyspace

db0:keys=958100,expires=934064,avg_ttl=2419035575
db10:keys=1,expires=0,avg_ttl=0

Replica:

Keyspace

db0:keys=958088,expires=934052,avg_ttl=2071806362
db10:keys=1,expires=0,avg_ttl=0
db13:keys=1,expires=1,avg_ttl=0

Okay, different number of keys could be explained by the time lag to switch between the consoles, but then db13 on the replica has some spare keys, atm there’s only one, but sometimes I get several thousands that seem to live for hours ! And the replication diff is like megabytes, which doesn’t happen in redis at all.

What am I doing wrong ?

Replica config:

127.0.0.1:6379> config get replicaof

  1. “replicaof”
  2. “10.100.34.2 6379”
    127.0.0.1:6379> config get slave-read-only
  3. “slave-read-only”
  4. “yes”

Furthermore - I don’t see the errors in the replica logs.