Does it make sense to have dynamic indexes in RediSearch? - redis

For example: consider that a hash (lets call it Event) has two searchable properties: user_id (num) and name (text).
However, whenever I need to filter events by name I aways have the user_id at hand.
So I am wondering if it makes sense to have one index of events per user instead of one big index of events for all users.
From my basic knowledge with Redis and RediSearch:
One index for all events of all users:
Prefix: "events:"
Key examples: events:123, events:456
Pros: easier.
Cons: whenever I need to search for events of name "foo" and user_id 100 RediSearch needs to find a block of events using the user_id and then filter the name. The hash needs to be in the same shard or use the coordinator.
One index per user for its events:
Prefixes: "events:%USER_ID%:", ie "events:789" where 789 is an user id
Key examples: events:789:123
Pros: Smaller indexes for better performance and can be easily distributed.
Cons: harder to maintain. If there are 1mi users we have 1mi indexes, dunno if this is a problem or not.

#jonathan it depends on your performance requirements and available memory.
If memory is not an issue (indexing the same doc twice) and you want to reduce the latency to a minimum then that sounds like an optional option.
With one caveat creating so many indexes has an overhead on the GC therefore you should only consider it if the user level index is pretty static or short live and you can define it as TEMPORARY (can be with endless timeout)
See: https://oss.redis.com/redisearch/Commands/#ftcreate

Related

Is it possible to do usual atomic INSERT operation but update Indexes asynchronously?

indexes make read fast but write slower. But why can't you have single writes and have db add indexes asynchronously with time, also cache in the INSERT until it's indexed?
Is there any database like that?
Converting my comments to an answer:
indexes make read fast but write slower
That's an oversimplification and it's also misleading.
Indexes make data lookups faster because the DBMS doesn't need to do a table-scan to find rows matching a predicate (the WHERE part of a query). Indexes don't make "reads" any faster (that's entirely dependent on the characteristics of your disk IO) and when used improperly they can sometimes even make queries slower (for reasons I won't get into).
I want to stress that the additional cost of writing to a single index, or even multiple indexes, when executing a DML statement (INSERT/UPDATE/DELETE/MERGE/etc) is negligible, really! (In actuality: foreign-key constraints are a much bigger culprit - and I note you can practically eliminate the cost of foreign-key constraint checking by adding additional indexes!). Indexes are primarily implemented using B-trees (a B-tree is essentially like a binary-tree, except rather than each node having only 2 children it can have many children because each tree-node comes with unused space for all those child node pointers, so inserting into the middle of a B-tree won't require data to be moved-around on-disk unlike with other kinds of trees, like a heap-tree).
Consider this QA where a Postgres user (like yourself) reports inserting 10,000 rows into a table. Without an index it took 78ms, with an index it took 84ms, that's only a 7.5% increase, which at that scale (6ms!) is so small it may as well be a rounding error or caused by IO scheduling. That should be proof enough it shouldn't be something you should worry about without actual hard data showing it's a problem for you and your application.
I assume you have this negative impression about indexes after reading an article like this one, which certainly gives the impression that "indexes are bad" - but while the points mentioned in that article are not wrong, there's a LOT of problems with that article so you shouldn't take it dogmatically. (I'll list my concerns with that article in the footer).
But why can't you have single writes and have db add indexes asynchronously with time
By this I assume you mean you'd like a DMBS to do a single-row INSERT by simply appending a new record to the end of a table and then immediately returning and then at an arbitrary point later the DBMS' housekeeping system would update the indexes afterwards.
The problem with that is that it breaks the A, C, and I parts of the the A.C.I.D. model.
Indexes are used for more than just avoiding table-scans: they're also used to store copies of table data for the benefit of queries that would use the index and which also need (for example) a small subset of the table's data, this significantly reduces disk reads. For this reason, RDBMS (and ISO SQL) allow indexes to include non-indexed data using the INCLUDES clause.
Consider this scenario:
CREATE INDEX IX_Owners ON cars ( ownerId ) INCLUDE ( colour );
CREATE INDEX IX_Names ON people ( name ) INCLUDE ( personId, hairColour );
GO;
SELECT
people.name,
people.hairColour,
cars.colour
FROM
cars
INNER JOIN people ON people.personId = cars.ownerId
WHERE
people.name LIKE 'Steve%'
The above query will not need to read either the cars or people tables on-disk. The DBMS will be able to fully answer the query using data only in the index - which is great because indexes tend to exist in a small number of pages on-disk which tend to be in proximal-locality which is good for performance because it means it will use sequential IO which scales much better than random IO.
The RDBMS will perform a string-prefix index-scan of the people.IX_Names index to get all of the personId (and hairColour) values, then it will look-up those personId values in the cars.IX_Owners index and be able to get the car.colour from the copy of the data inside the IX_Owners index without needing to read the tables directly.
Now, assuming that another database client has just completed inserted a load of records into the cars and/or people table with a COMMIT TRANSACTION just for good measure, and the RDMBS uses your idea of only updating indexes later whenever it feels like it, then if that same database client re-runs the query from above it would return stale data (i.e. wrong data) because the query uses the index, but the index is old.
In addition to using index tree nodes to store copies of table data to avoid non-proximal disk IO, many RDBMS also use index-trees to store entire copies - even multiple copies of table data, to enable other scenarios, such as columnar data storage and indexed-VIEWs - both of these features absolutely require that indexes are updated atomically with table data.
Is there any database like that?
Yes, they exist - but they're not widely used (or they're niche) because for the vast majority of applications it's entirely undesirable behaviour for the reasons described above.
There are distributed databases that are designed around eventual consistency, but clients (and entire application code) needs to be designed with that in-mind, and it's a huge PITA to have to redesign a data-centric application to support eventual-consistency which is why you only really see them being used in truly massive systems (like Facebook, Google, etc) where availability (uptime) is more important than users seeing stale-data for a few minutes.
Footnote:
Regarding this article: https://use-the-index-luke.com/sql/dml/insert
The number of indexes on a table is the most dominant factor for insert performance. The more indexes a table has, the slower the execution becomes. The insert statement is the only operation that cannot directly benefit from indexing because it has no where clause.
I disagree. I'd argue that foreign-key constraints (and triggers) are far more likely to have a larger detrimental effect on DML operations.
Adding a new row to a table involves several steps. First, the database must find a place to store the row. For a regular heap table—which has no particular row order—the database can take any table block that has enough free space. This is a very simple and quick process, mostly executed in main memory. All the database has to do afterwards is to add the new entry to the respective data block.
I agree with this.
If there are indexes on the table, the database must make sure the new entry is also found via these indexes. For this reason it has to add the new entry to each and every index on that table. The number of indexes is therefore a multiplier for the cost of an insert statement.
This is true, but I don't know if I agree that it's a "multiplier" of the cost of an insert.
For example, consider a table with hundreds of nvarchar(1000) columns and several int columns - and there's separate indexes for each int column (with no INCLUDE columns). If you're inserting 100x megabyte-sized rows all-at-once (using an INSERT INTO ... SELECT FROM statement) the cost of updating those int indexes is very likely to require much less IO than the table data.
Moreover, adding an entry to an index is much more expensive than inserting one into a heap structure because the database has to keep the index order and tree balance. That means the new entry cannot be written to any block—it belongs to a specific leaf node. Although the database uses the index tree itself to find the correct leaf node, it still has to read a few index blocks for the tree traversal.
I strongly disagree with this, especially the first sentence: "adding an entry to an index is much more expensive than inserting one into a heap structure".
Indexes in RDBMS today are invariably based on B-trees, not binary-trees or heap-trees. B-trees are essentially like binary-trees except each node has built-in space for dozens of child node pointers and B-trees are only rebalanced when a node fills its internal child pointer list, so a B-tree node insert will be considerably cheaper than the article is saying because each node will have plenty of empty space for a new insertion without needing to re-balance itself or any other relatively expensive operation (besides, DBMS can and do index maintenance separately and independently of any DML statement).
The article is correct about how the DBMS will need to traverse the B-tree to find the node to insert into, but index nodes are efficently arranged on-disk, such as keeping related nodes in the same disk page which minimizes index IO reads (assuming they aren't already loaded into memory first). If an index tree is too big to store in-memory the RDBMS can always keep a "meta-indexes" in-memory so it could potentially instantly find the correct B-tree index without needing to traverse the B-tree from the root.
Once the correct leaf node has been identified, the database confirms that there is enough free space left in this node. If not, the database splits the leaf node and distributes the entries between the old and a new node. This process also affects the reference in the corresponding branch node as that must be duplicated as well. Needless to say, the branch node can run out of space as well so it might have to be split too. In the worst case, the database has to split all nodes up to the root node. This is the only case in which the tree gains an additional layer and grows in depth.
In practice this isn't a problem, because the RDBMS's index maintenance will ensure there's sufficient free space in each index node.
The index maintenance is, after all, the most expensive part of the insert operation. That is also visible in Figure 8.1, “Insert Performance by Number of Indexes”: the execution time is hardly visible if the table does not have any indexes. Nevertheless, adding a single index is enough to increase the execute time by a factor of a hundred. Each additional index slows the execution down further.
I feel the article is being dishonest by suggesting (implying? stating?) that index-maintenance happens with every DML. This is not true. This may have been the case with some early dBase-era databases, but this is certainly not the case with modern RDBMS like Postgres, MS SQL Server, Oracle and others.
Considering insert statements only, it would be best to avoid indexes entirely—this yields by far the best insert performance.
Again, this claim in the article is not wrong, but it's basically saying if you want a clean and tidy house you should get rid of all of your possessions. Indexes are a fact of life.
However tables without indexes are rather unrealistic in real world applications. You usually want to retrieve the stored data again so that you need indexes to improve query speed. Even write-only log tables often have a primary key and a respective index.
Indeed.
Nevertheless, the performance without indexes is so good that it can make sense to temporarily drop all indexes while loading large amounts of data—provided the indexes are not needed by any other SQL statements in the meantime. This can unleash a dramatic speed-up which is visible in the chart and is, in fact, a common practice in data warehouses.
Again, with modern RDBMS this isn't necessary. If you do a batch insert then a RDBMS won't update indexes until after the table-data has finished being modified, as a batch index update is cheaper than many individual updates. Similarly I expect that multiple DML statements and queries inside an explicit BEGIN TRANSACTION may cause an index-update deferral provided no subsequent query in the transaction relies on an updated index.
But my biggest issue with that article is that the author is making these bold claims about detrimental IO performance without providing any citations or even benchmarks they've run themselves. It's even more galling that they posted a bar-chart with arbitrary numbers on, again, without any citation or raw benchmark data and instructions for how to reproduce their results. Always demand citations and evidence from anything you read making claims: because the only claims anyone should accept without evidence are logical axioms - and a quantitative claim about database index IO cost is not a logical axiom :)
For PostgreSQL GIN indexes, there is the fastupdate feature. This stores new index entries into a unordered unconsolidated area waiting for some other process to file them away into the main index structure. But this doesn't directly match up with what you want. It is mostly designed so that the index updates are done in bulk (which can be more IO efficient), rather than in the background. Once the unconsolidated area gets large enough, then a foreground process might take on the task of filing them away, and it can be hard to tune the settings in a way to get this to always be done by a background process instead of a foreground process. And it only applies to GIN indexes. (With the use of the btree_gin extension, you can create GIN indexes on regular scalar columns rather than the array-like columns it usually works with.) While waiting for the entries to be consolidated, every query will have to sequential scan the unconsolidated buffer area, so delaying the updates for the sake of INSERT can come at a high cost for SELECTs.
There are more general techniques to do something like this, such as fractal tree indexes. But these are not implemented in PostgreSQL, and wherever they are implemented they seem to be proprietary.

Performance difference in Couchbase's get by Key and select by index

As we are doing benchmark tests on our Couchbase DB, we tried to compare search for item by their id / key and search for items by a query that uses secondary index.
Following this article about indexing and performance in Couchbase we thought the performance of the two will be the same.
However, in our tests, we discovered that sometimes, the search by key/id was much faster then the search that uses the secondary index.
E.g. ~3MS to search using the index and ~0.3MS to search by the key.(this is a 10 times factor)
The point is that this difference is not consist. The search by key varies from 0.3MS to 15MS.
We are wondering if:
There should be better performance for search by key over search by secondary index?
There should be such time difference between key searches?
The results you get are consistent with what I would expect. Couchbase works as a key-value store when you do any operation using the id. A key-value store is roughly a big distributed hashmap, and in this data structure, you can a very good performance on get/save/delete while using the id.
Whenever you store a new document, couchbase hash the key and assign a Virtual Bucket to it (something similar to a shard). When you need to get this document back, it uses the same algorithm to find out in which virtual bucket the document is located, as the SDK has the cluster map and knows exactly which node has which shards, your application will request the document directly to the node who owns it.
On the other hand, when you query the database, Couchbase has to make internally a map/reduce to find out where the document is located, that is why operations by id are faster.
About your questions about results from 0.3ms to 15ms, it is hard to tell without debugging your environment. However, there are a number of factors that could contribute to it. Ex: the document is cached/not cached, node is undersized, etc.
To add to #deniswrosa's answer, the secondary index will always be slower, because first the index must be traversed based on your query to find the document key, and then a key lookup is performed. Doing just the key lookup is faster if you already have the key. The amount of work to traverse the index can vary depending on how selective the index is, whether the entire index is in memory, etc. Memory-optimized indexes can ensure that the whole index is in memory, if you have enough memory to support that.
Of course even a simple key lookup can be slower if the document in question is not in the cache, and needs to be brought in to memory from storage.
It is possible to achieve sub-millisecond secondary lookups at scale, but it requires some tuning of your query, index, and some possibly some of Couchbase' system parameters. Consider the following simple example:
Sample document in userBucket:
"user::000000000001" : {
"email" : "benjamin1#couchbase.com",
"userId" : "000000000001"
}
This query:
SELECT userId
FROM userBucket
WHERE
email = "benjamin1#couchbase.com"
AND userId IS NOT NULL
;
...should be able to achieve sub-millisecond performance with a properly tuned secondary index:
CREATE INDEX idx01 ON userBucket(email, userId);
Since the index is covering the query completely there is no need for the Query engine to FETCH the document from the K/V store. However "SELECT * ..." will always cause the Query service to FETCH the document and thus will be slower than a simple k/v GET("user::000000000001").
For the best latencies, make sure to review your query plan (using EXPLAIN syntax) and make sure your query is not FETCHing. https://docs.couchbase.com/server/6.0/n1ql/n1ql-language-reference/explain.html

Is using a timestamp as a hash key on a GSI in DynamoDB a good approach

I have a large (2B + records) DynamoDB table.
I want to implement a distributed locking process by adding a new field, 'index_due_at' when an item is created or updated. After the create/update, I will do some further processing on the item and then remove the 'index_due_at' field.
I'd like to create a sweeper job which will periodically extract any records with an outstanding 'index_due_at' field (on the assumption that something about the above process failed) to give those records further treatment. I would anticipate at most 100s of records in this state at any one time, more likely 10s.
To optimise the performance of the sweeper, I want to create a GSI including the new field (and project the key data into it).
It seems that using a timestamp (in millis) as the GSI HASH key ought to give a good distribution. And I don't need to query on this field's value, just on its presence. Can anyone identify any drawbacks in this approach and if so, suggest an alternative?
Issues I can anticipate include:
* Non-uniqueness in timestamps at milli level.
* Possible hash key problems with numeric values?
* Possible hash key problems with numeric values that don't vary much in the most significant digits.
This is less of a problem than you might be thinking. GSI hash keys don't actually have to be unique, so you're fine on than front.
You probably already know this, but your GSI will only contain items with GSI keys, so your GSI should be pretty small (100s of items).
One thought I have is that the index_due_at might actually be better as a GSI sort key rather than hash key. Data is sorted within a partition by sort key. So you could have a GSI hash key of index_due_at_flag which would be Y if present, then a sort key of index_due_at. This would mean all your data would be sorted naturally, so you could process it in date order.
That said, you are probably never going to Query this GSI, so I suspect your choice of keys hardly matters at all. Presumably you will just do a Scan, get all the items and try and process them all. In which case you would never even use the keys. Just having a key attribute present would put the item in the GSI.
Another thought is that you need to handle the fact GSIs are not perfectly synchronous with the base table. Its possible (admittedly unlikely) that an item in your GSI has actually just been processed. Therefore if your sweeper script picks up an item from the GSI, you should handle the fact its possible its already been updated in the base table (e.g. by checking the base table item before attempting to process it).
Good luck with it. I answered because I liked your bio! Hope staying on the right side of barrel shaped is working out :)
This should be a perfect scenario for using DynamoDB Sparse Index
Use the 'index_due_at' as sort key in GSI, and only the items you are interested will be in the index, greatly reducing the space needed and the performance.

Out of Process in memory database table that supports queries for high speed caching

I have a SQL table that is accessed continually but changes very rarely.
The Table is partitioned by UserID and each user has many records in the table.
I want to save database resources and move this table closer to the application in some kind of memory cache.
In process caching is too memory intensive so it needs to be external to the application.
Key Value stores like Redis are proving inefficient due to the overhead of serializing and deserializing the table to and from Redis.
I am looking for something that can store this table (or partitions of data) in memory, but let me query only the information I need without serializing and deserializing large blocks of data for each read.
Is there anything that would provide Out of Process in memory database table that supports queries for high speed caching?
Searching has shown that Apache Ignite might be a possible option, but I am looking for more informed suggestions.
Since it's out-of-process, it has to do serialization and deserialization. The problem you concern is how to reduce the serialization/deserizliation work. If you use Redis' STRING type, you CANNOT reduce these work.
However, You can use HASH to solve the problem: mapping your SQL table to a HASH.
Suppose you have the following table: person: id(varchar), name(varchar), age(int), you can take person id as key, and take name and age as fields. When you want to search someone's name, you only need to get the name field (HGET person-id name), other fields won't be deserialzed.
Ignite is indeed a possible solution for you since you may optimize serialization/deserialization overhead by using internal binary representation for accessing objects' fields. You may refer to this documentation page for more information: https://apacheignite.readme.io/docs/binary-marshaller
Also access overhead may be optimized by disabling copy-on-read option https://apacheignite.readme.io/docs/performance-tips#section-do-not-copy-value-on-read
Data collocation by user id is also possible with Ignite: https://apacheignite.readme.io/docs/affinity-collocation
As the #for_stack said, Hash will be very suitable for your case.
you said that Each user has many rows in db indexed by the user_id and tag_id . So It is that (user_id, tag_id) uniquely specify one row. Every row is functional depends on this tuple, you could use the tuple as the HASH KEY.
For example, if you want save the row (user_id, tag_id, username, age) which values are ("123456", "FDSA", "gsz", 20) into redis, You could do this:
HMSET 123456:FDSA username "gsz" age 30
When you want to query the username with the user_id and tag_id, you could do like this:
HGET 123456:FDSA username
So Every Hash Key will be a combination of user_id and tag_id, if you want the key to be more human readable, you could add a prefix string such as "USERINFO". e.g. : USERINFO:123456:FDSA .
BUT If you want to query with only a user_id and get all rows with this user_id, this method above will be not enough.
And you could build the secondary indexes in redis for you HASH.
as the above said, we use the user_id:tag_id as the HASH key. Because it can unique points to one row. If we want to query all the rows about one user_id.
We could use sorted set to build a secondary indexing to index which Hashes store the info about this user_id.
We could add this in SortedSet:
ZADD user_index 0 123456:FDSA
As above, we set the member to the string of HASH key, and set the score to 0. And the rule is that we should set all score in this zset to 0 and then we could use the lexicographical order to do range query. refer zrangebylex.
E.g. We want to get the all rows about user_id 123456,
ZRANGEBYLEX user_index [123456 (123457
It will return all the HASH key whose prefix are 123456, and then we use this string as HASH key and hget or hmget to retrieve infomation what we want.
[ means inclusive, and ( means exclusive. and why we use 123457? it is obvious. So when we want to get all rows with a user_id, we shoud specify the upper bound to make the user_id string's leftmost char's ascii value plus 1.
More about lex index you could refer the article I mentioned above.
You can try apache mnemonic started by intel. Link -http://incubator.apache.org/projects/mnemonic.html. It supports serdeless features
For a read-dominant workload MySQL MEMORY engine should work fine (writing DMLs lock whole table). This way you don't need to change you data retrieval logic.
Alternatively, if you're okay with changing data retrieval logic, then Redis is also an option. To add to what #GuangshengZuo has described, there's ReJSON Redis dynamically loadable module (for Redis 4+) which implements document-store on top of Redis. It can further relax requirements for marshalling big structures back and forth over the network.
With just 6 principles (which I collected here), it is very easy for a SQL minded person to adapt herself to Redis approach. Briefly they are:
The most important thing is that, don't be afraid to generate lots of key-value pairs. So feel free to store each row of the table in a different key.
Use Redis' hash map data type
Form key name from primary key values of the table by a separator (such as ":")
Store the remaining fields as a hash
When you want to query a single row, directly form the key and retrieve its results
When you want to query a range, use wild char "*" towards your key. But please be aware, scanning keys interrupt other Redis processes. So use this method if you really have to.
The link just gives a simple table example and how to model it in Redis. Following those 6 principles you can continue to think like you do for normal tables. (Of course without some not-so-relevant concepts as CRUD, constraints, relations, etc.)
using Memcache and REDIS combination on top of MYSQL comes to Mind.

How to implement a scalable, unordered collection in DynamoDB?

I am looking into implementing a scalable unordered collection of objects on top of Amazon DynamoDB. So far the following options have been considered:
Use DynamoDB document data types (map, list) and use document path to access stand-alone items. This has one obvious drawback for collection being limited to 400KB of data, meaning perhaps 1..10K objects depending on their size. Less obvious drawback is that cost of insertion of a new object into such collection is going to be huge: Amazon specifies that the write capacity will be deducted based on the total item size, not just newly added object -- therefore ~400 capacity units for inserting 1KB object when approaching the size limit. So considering this ruled out?
Using composite primary hash + range key, where primary hash remains the same for all objects in the collection, and range key is just something random or an atomic counter. Obvious drawback is that having identical hash key results in bad key distribution -- cardinality is low when there are collections with large number of objects. This means bad partitioning, and having a scale issue with all reads/writes on the same collection being stuck to one shard, becoming subject to 3000 reads / 1000 writes per second limitation of DynamoDB partition.
Using global secondary index with secondary hash + range key, where hash key remains the same for all objects belonging to the same collection, and range key is just something random or an atomic counter. Similar to above, partitioning becomes poor for the GSI, and it will become a bottleneck with too many identical hashes draining all the provisioned capacity to the index rapidly. I didn't find how the GSI is implemented exactly, thus not sure how badly it suffers from low cardinality.
Question is, whether I could live with (2) or (3) and suffer from non-ideal key distribution, or is there another way of implementing collection that was overlooked, or perhaps I should at all consider looking into another nosql database engine.
This is a "shooting from the hip" answer, what you end up doing may depend on how much and what type of reading and writing you do.
Two things the dynamo docs encourage you to avoid are hot keys and, in general, scans. You noted that in cases (2) and (3), you end up with a hot key. If you expect this to scale (large collections), the hot key will probably hurt more and more, especially if this is a write-intensive application.
The docs on Query and Scan operations (http://docs.aws.amazon.com/amazondynamodb/latest/developerguide/QueryAndScan.html) say that, for a query, "you must specify the hash key attribute name and value as an equality condition." So if you want to avoid scans, this might still force your hand and put you back into that hot key situation.
Maybe one route would be to embrace doing a scan operation, but just have one table devoted to your collection. Then you could just have a fully random (well distributed) hash key and do a scan every time. This assumes you always want everything from the collection (you didn't say). This will still hurt if you scale up to a large collection, but if you always want the full set back, you'll have to deal with that pain regardless. If you just want a subset, you can add a limit parameter. This would help performance, but you will always get back the same subset (or you can use the last evaluated key and keep going). The docs also mention parallel scans.
If you are using AWS, elasticache/redis might be another route to try? The first pass might code up a lot faster/cleaner than situation (1) that you mentioned.