Squirro uses caches to speed up common actions, such as queries. This section shows how to flush these caches when necessary.
Table of Contents
Introduction
In rare circumstances, the Squirro caches may become stale. If and when this happens, please involve Squirro Support, as this is considered a bug.
Caches
There are two different caches managed in Redis. They are stored in different databases (seeĀ Accessing Redis Server for background):
- db7: the query cache, which stores the full query responses. Updating this cache can be disabled by using theĀ
update_cache
option in the item query API call. - db9: the generic HTTP cache which is used to cache internal calls to the various internal web services.
Flush Cache (Squirro 2.4.6 and newer)
To flush the caches, restart redis-server-cache:
# In the case of CentOS 6 / RHEL 6, run: monit restart redis-server-cache # In the case of CentOS 7, run: systemctl restart redis-server-cache
Flush Cache (Squirro 2.4.5 and older)
To flush a cache, connect to Redis and specify a password, then select the right database, and then flush the cache.
Beware, that flushing the incorrect database can affect the operation of Squirro. So please double-check that the right database is selected before flushing.
Below is an example session to flush the query cache (database 7).
[squirro@localhost ~]$ redis-cli 127.0.0.1:6379> auth REDIS_PASSWORD OK 127.0.0.1:6379> select 7 OK 127.0.0.1:6379[7]> dbsize (integer) 114 127.0.0.1:6379[7]> flushdb OK 127.0.0.1:6379[7]> dbsize (integer) 0