In redis I'm planning to store key as a unique string and value will be a list.
I have a use case where I need to do 2 things.
First, I need to get all the values associated with a key by providing the key as input.
Second, I want to get all the keys associated with a value by providing one of the value in the values list.
Second part is where I need the advice, how we can achive this ?
I cannot get all the keys or key value pair and loop through because I will have millions of entries in Redis.
As mentioned in the comment above the retrieving of all keys with associated value at will probably sometimes create a performance issue as this will be a run through large entries.As also suggested in the official documentation about retrieving data from the memory caches you can try and use the following Redis command to get the value and see if that is what can solve your purpose.
GET
MGET
Related
I am implementing a feed for a social network in that newly uploaded post should be served first and so on.I use hashes of posts as keys and posts as values.I need the posts in "newest first order".How to do it?
My idea is
Store the post and timestamp with hash as a key
Get all keys and timestamps
Sort the timestamps in descending order
Then use the respective keys to get th latest images
Question1:But this approach is not good.How to do it ?
EDIT:Question2:Please tell me what algorithm you use to serve feed.If feed is common for all users based on "newest-first",how to implement it?
This is my first time in backend.Please if the question is dumb.
Thanks.
Here are three options for you:
Use a sorted set, using the timestamp as score, and the post-hash as value. The post-hash is also the key in a hash where the actual posts are stored. Commands involved: ZADD, HSET, ZREVRANGEBYSCORE, HGET.
Use a sorted set, using the timestamp as score, and the post with metadata as value. Make sure "post with metadata" is unique, you can include the timestamp and user to achieve this. This will have better performance, but makes it a bit harder if you have to find a specific post. Commands involved: ZADD, ZREVRANGEBYSCORE, ZRANGEBYSCORE.
Use Redis Streams. If you want a uniform insert order independent of client time, Redis can set the timestamp for you. However, stream entries cannot be modified, so either users cannot edit posts, or whenever they edit the post is brought up as new. Commands involved: XADD, XREVRANGE, XDEL.
See:
Redis Commands
Introduction to Redis Streams
I currently have a scenario where we are using REDIS to store string field-value pairs within a hashed set HSET.
The original reasoning behind using hashed sets instead of just sets is ease of retrieving the records using HSCAN inside a GUI Search Bar as opposed to just SCAN because it's easier to get the length of a hash to use in the COUNT field.
I read in the Redis documentation that both GET and HGET commands execute with O(1) time complexity, but a member of my team thinks that if I store all the values inside a single key then it basically returns the entire hash during HGET instead of the singular field-value that I need.
So for a made up but similar example:
I have a Redis instance with a single Hashed Set called users.
The hashed set has 150,000 field:value pairs of username:email
If when I execute hget users coolguy, is the entire hash getting returned or just the email for user coolguy?
First of all, HSET is not a hash set, it creates a hash table. The mechanism behind the hash table and set (which is indeed a hash set) in redis is the same, the difference is mainly that the hash table has values.
To answer your question:
If when I execute hget users coolguy, is the entire hash getting returned or just the email for user coolguy?
Just the email for that user. You can also use HMGET to get the emails of multiple users at once. It's O(1) for each user you fetch, or O(n) for n users.
I have an map function in a view in CouchDB that emits non-unique two array keys, for documents of type message, e.g.
The first position in the array key is a user_id, the second position represents whether or not the user has read the message.
This works nicely in that I can set include_docs=true and retrieve the actual documents. However, I'm retrieving duplicate documents in that case, as you can see above in the view results. I need to be able to write a view that can be queried to return unique messages that have been read by a given user. Additionally, I need to be able to efficiently paginate the resultset.
notice in the image above that [66, true] is emitted twice for doc id 26a9a271de3aac494d37b17334aaf7f3. As far as I can tell, with the keys in my map function, I cannot reduce in such a way that unique documents will be returned.
the next idea I had was to emit doc._id also in the map function and reduce with group_level=exact the result being:
now I am able to get unique document ids, but I cannot get the documents without doing a second query. And even in the case of a second query, it will require a lot of complexity to do pagination like this (at least I think so).
the last idea I came up with is to emit the entire document rather than the doc._id in the third position in the array key, then I can access the entire document and likely paginate. This seems really brutish.
So my question is:
Is #3 above a terrible idea? Is there something I'm missing? Is there a better approach?
Thanks in advance.
See #WickedGrey's comment to the question. The solution is to ensure that I never emit the same key twice for one document. I do this in the map function by keeping track of the keys as I emit them in an array, then skipping the emit if the key exists in the array.
Let's assume I have a keyspace with a column family that stores user objects and the key of these objects is the username.
How can I use Hector to get a list of users sorted by username?
I tried to use a RangeSlicesQuery, paging works fine with this query, but the results are not sorted in any way.
I'm an absolute Cassandra beginner, can anyone point me to a simple example that shows how to sort a column family by key? Please ask if you need more details on my efforts.
Edit:
The result was not sorted because I used the default RandomPartitioner instead of the OrderPreseveringPartitioner in cassandra.yaml.
Probably it's better not to rely on the sorting by key but to use a secondary index.
Quoting Cassandra - The Definitive Guide
Column names are stored in sorted order according to the value of compare_with. Rows,
on the other hand, are stored in an order defined by the partitioner (for example,
with RandomPartitioner, they are in random order, etc.)
I guess you are using RandomPartitioner which
... return data in an essentially random order.
You should probably use OrderPreservingPartitioner (OPP) where
Rows are therefore stored
by key order, aligning the physical structure of the data with your sort order.
Be aware of inefficiency of OPP.
(edit on Mar 07, 2014)
Important:
This answer is very old now.
It is a system-wide setting. You can set in cassandra.yaml. See this doc. Again, OPP is highly discouraged. This document is for version 1.1, and you can see it is deprecated. It is likely that it is removed from latest version. If you do want to use OPP, you may want to revisit the architecture the architecture.
Or create a row called "meta:userNames" in same column family and put all user names as a look up hash. Something like that.
Users {
key: "meta:userNames" {david:david, paolo:paolo, victor:victor},
key: "paolo" {password:"*****", locale:"it_it"},
key: "david" {password:"*****", locale:"en_us"},
key: "victor" {password:"*****", locale:"en_uk"}
}
First query the meta:userNames columns (that are sorted) and use them to get the user rows. Don't try to get everything via single db query as in SQL driven databases. Use Cassandra as huge Hash Map which provides rapid random access to its data.
Is it currently only possible to expire an entire key/value pair? What if I want to add values to a List type structure and have them get auto removed 1 hour after insertion. Is that currently possible, or would it require running a cron job to do the purging manually?
There is a common pattern that solves this problem quite well.
Use sorted sets, and use a timestamp as the score. It's then trivial to delete items by score range, which could be done periodically, or only on every write, with reads always ignoring the out of range elements, by reading only a range of scores.
More here: https://groups.google.com/forum/#!topic/redis-db/rXXMCLNkNSs
Is it currently only possible to expire an entire key/value pair?
As far as I know, and also according to key commands and document about expiration, currently you can set expiration only to specific key and not to it's underlying data structure. However there is a discussion on google groups about this functionality with outlined alternative solutions.
I came upon a different method of handling this, don't know if it's helpful to any of you,but here goes:
The hash and the sorted set are linked by a guid.
I have a hash that is set to expire in 'x' seconds
I have a sorted set that is used for ranged queries
The data for both is added in a transaction, so if one fails, they both fail.
Upon a ranged query, use 'EXISTS' to see if the hashed value exists as the results are iterated over
If it does not exist, it has expired, so delete the item from the sorted set
What about creating two seperate sorted sets?
Main sorted set which is key = value.
Expire sorted set which is key = expire_timestamp.
If you only want to expire a single score you can set as key:unique_id = expire_timestamp.
With mercy of zrangebyscore we can get expired keys. Then all we need to do is check periodically and zrem.
If you only want to expire a single score: ​zincrby -1.