When using `Discourse.cache.fetch` with an expiry, there's a potential for a race condition due to how we read the data from redis.
The code used to be
```ruby
raw = redis.get(key) if !force
entry = read_entry(key) if raw
return entry if raw && !(entry == :__corrupt_cache__)
```
with `read_entry` defined as follow
```ruby
def read_entry(key)
if data = redis.get(key)
Marshal.load(data)
end
rescue => e
:__corrupt_cache__
end
```
If the value at "key" expired in redis between `raw = redis.get` and `entry = read_entry`, the `entry` variable would be `nil` despite `raw` having a value.
We would then proceed to return `entry` (which is `nil`) thinking it had a value, when it didn't.
The first `redis.get` can be skipped altogether and we can rely only on `read_entry` to read the data from redis. Thus avoiding the race condition and removing the double read operations.
Internal ref - t/132507
Currently when a cache entry is corrupt, we log the event without doing
anything else. It means the cache is still corrupt, and the proper value
isn’t computed again.
Normally, it’s very rare the cache becomes corrupt, but it can happen
when upgrading Rails for example and the cache format changes. This is
normally handled automatically by Rails but since we’re using a custom
cache class, we have to do it ourselves.
This patch takes the same approach the Rails team did, when a cache
entry is corrupt, we treat it as a miss, recomputing the proper value
and caching it in the new format.
This is a bottom up rewrite of Discourse cache to support faster performance
and a limited surface area.
ActiveSupport::Cache::Store accepts many options we do not use, this partial
implementation only picks the bits out that we do use and want to support.
Additionally params are named which avoids typos such as "expires_at" vs "expires_in"
This also moves a few spots in Discourse to use Discourse.cache over setex
Performance of setex and Discourse.cache.write is similar.
This reduces chances of errors where consumers of strings mutate inputs
and reduces memory usage of the app.
Test suite passes now, but there may be some stuff left, so we will run
a few sites on a branch prior to merging
As per the documentation for KEYS
```
Warning: consider KEYS as a command that should only be used in production environments with extreme care. It may ruin performance when it is executed against large databases. This command is intended for debugging and special operations, such as changing your keyspace layout.
```
Instead SCAN
```
Since these commands allow for incremental iteration, returning only a small number of elements per call, they can be used in production without the downside of commands like KEYS or SMEMBERS that may block the server for a long time (even several seconds) when called against big collections of keys or elements.
```
implemented an ActiveSupport::Cache::Store for our internal use.
* allows for expire by family
* works correctly in multisite
* namespaced correctly
Removed redis-rails from the project, no longer needed
Régis Hanol
Loïc Guitaut
David Taylor
Dan Ungureanu
Joffrey JAFFEUX
Guo Xiang Tan
Sam Saffron
Guo Xiang Tan