Efficiently update redis by member value

Asked Oct 29 '18 at 02:23

Active Oct 29 '18 at 02:23

Viewed 46 times

Let's suppose I have the following three k/v pairs:

user+123 = {
   'name': 'tom',
   'company': 'nike'
   'status': 0
}

user+199 = {
   'name': 'sarah',
   'company': 'nike'
   'status': 0
}

user+359 = {
   'name': 'brad',
   'company': 'nike'
   'status': 0
}

But suppose there may be any number of these user ids (actually, there would be various other object types, such as users, companies, documents, etc.) and it could be many million.

What would be the most efficient to do the following update:

update user* set status=1 where company=nike

For example, to do it for one object I could do:

r.set('user+123') = {
   'name': 'tom',
   'company': 'nike'
   'status': 1 // changed here
}

Perhaps a solution showing how to use redis search would be the best approach here? As opposed to having to iterate through all objects and check the values of the objects.

asked Oct 29 '18 at 02:23

1

Seems duplicate with [this question](https://stackoverflow.com/questions/53028596/store-user-information-with-redis-or-mongodb/53043377#53043377) – for_stack Oct 29 '18 at 10:22
1

You need a secondary index. RediSearch is the best one available ATM. (disclaimer - I wrote it but no longer maintaining it, other people are). – Not_a_Golfer Oct 29 '18 at 11:41
What they both said. – Itamar Haber Oct 29 '18 at 13:19

Efficiently update redis by member value

0 Answers0