How to combine search text with other criteria using Redis? - lua

Question

I successfully wrote an intersection of text search and other criteria using Redis. To achieve that I'm using a Lua script. The issue is that I'm not only reading, but also writing values from that script. From Redis 3.2 it's possible to achieve that by calling redis.replicate_commands(), but not before 3.2.
Below is how I'm storing the values.
Names
> HSET product:name 'Cool product' 1
> HSET product:name 'Nice product' 2
Price
> ZADD product:price 49.90 1
> ZADD product:price 54.90 2
Then, to get all products that matches 'ice', for example, I call:
> HSCAN product:name 0 MATCH *ice*
However, since HSCAN uses a cursor, I have to call it multiple times to fetch all results. This is where I'm using a Lua script:
local cursor = 0
local fields = {}
local ids = {}
local key = 'product:name'
local value = '*' .. ARGV[1] .. '*'
repeat
local result = redis.call('HSCAN', key, cursor, 'MATCH', value)
cursor = tonumber(result[1])
fields = result[2]
for i, id in ipairs(fields) do
if i % 2 == 0 then
ids[#ids + 1] = id
end
end
until cursor == 0
return ids
Since it's not possible to use the result of a script with another call, like SADD key EVAL(SHA) .... And also, it's not possible to use global variables within scripts. I've changed the part inside the fields' loop to access the list of ID's outside the script:
if i % 2 == 0 then
ids[#ids + 1] = id
redis.call('SADD', KEYS[1], id)
end
I had to add redis.replicate_commands() to the first line. With this change I can get all ID's from the key I passed when calling the script (see KEYS[1]).
And, finally, to get a list 100 product ID's priced between 40 and 50 where the name contains "ice", I do the following:
> ZUNIONSTORE tmp:price 1 product:price WEIGHTS 1
> ZREMRANGEBYSCORE tmp:price 0 40
> ZREMRANGEBYSCORE tmp:price 50 +INF
> EVALSHA b81c2b... 1 tmp:name ice
> ZINTERSTORE tmp:result tmp:price tmp:name
> ZCOUNT tmp:result -INF +INF
> ZRANGE tmp:result 0 100
I use the ZCOUNT call to know in advance how many result pages I'll have, doing count / 100.
As I said before, this works nicely with Redis 3.2. But when I tried to run the code at AWS, which only supports Redis up to 2.8, I couldn't make it work anymore. I'm not sure how to iterate with HSCAN cursor without using a script or without writing from the script. There is a way to make it work on Redis 2.8?
Some considerations:
I know I can do part of the processing outside Redis (like iterate the cursor or intersect the matches), but it'll affect the application overall performance.
I don't want to deploy a Redis instance by my own to use version 3.2.
The criteria above (price range and name) is just an example to keep things simple here. I have other fields and type of matches, not only those.
I'm not sure if the way I'm storing the data is the best way. I'm willing to listen suggestion about it.

Answer 1

The only problem I found here is storing the values inside a lua scirpt. So instead of storing them inside a lua, take that value outside lua (return that values of string[]). Store them in a set in a different call using sadd (key,members[]). Then proceed with intersection and returning results.
> ZUNIONSTORE tmp:price 1 product:price WEIGHTS 1
> ZREVRANGEBYSCORE tmp:price 0 40
> ZREVRANGEBYSCORE tmp:price 50 +INF
> nameSet[] = EVALSHA b81c2b... 1 ice
> SADD tmp:name nameSet
> ZINTERSTORE tmp:result tmp:price tmp:name
> ZCOUNT tmp:result -INF +INF
> ZRANGE tmp:result 0 100
IMO your design is the most optimal one. One advice would be to use pipeline wherever possible, as it would process everything at one go.
Hope this helps
UPDATE
There is no such thing like array ([ ]) in lua you have to use the lua table to achieve it. In your script you are returning ids right, that itself is an array you can use it as a separate call to achieve the sadd.
String [] nameSet = (String[]) evalsha b81c2b... 1 ice -> This is in java
SADD tmp:name nameSet
And the corresponding lua script is the same as that of your 1st one.
local cursor = 0
local fields = {}
local ids = {}
local key = 'product:name'
local value = '*' .. ARGV[1] .. '*'
repeat
local result = redis.call('HSCAN', key, cursor, 'MATCH', value)
cursor = tonumber(result[1])
fields = result[2]
for i, id in ipairs(fields) do
if i % 2 == 0 then
ids[#ids + 1] = id
end
end
until cursor == 0
return ids

Answer 2

The problem isn't that you're writing to the database, it's that you're doing a write after a HSCAN, which is a non-deterministic command.
In my opinion there's rarely a good reason to use a SCAN command in a Lua script. The main purpose of the command is to allow you to do things in small batches so you don't lock up the server processing a huge key space (or hash key space). Since scripts are atomic, though, using HSCAN doesn't help—you're still locking up the server until the whole thing's done.
Here are the options I can see:
If you can't risk locking up the server with a lengthy command:
Use HSCAN on the client. This is the safest option, but also the slowest.
If you're want to do as much processing in a single atomic Lua command as possible:
Use Redis 3.2 and script effects replication.
Do the scanning in the script, but return the values to the client and initiate the write from there. (That is, Karthikeyan Gopall's answer.)
Instead of HSCAN, do an HKEYS in the script and filter the results using Lua's pattern matching. Since HKEYS is deterministic you won't have a problem with the subsequent write. The downside, of course, is that you have to read in all of the keys first, regardless of whether they match your pattern. (Though HSCAN is also O(N) in the size of the hash.)