Is it suboptimal to add the same datoms multiple times? - clojure

I'm currently using Datomic in one of my project, and a question is bothering me.
Here is a simplified version of my problem:
I need to parse a list of small English sentences and insert both the full sentence and its words into Datomic.
the file that contains the list of sentences is quite big (> 10 GB)
the same sentence can occur multiple times in the file, and their words can also occur multiple times across sentences
during the insertion process, an attribute will set to associate each sentence with its corresponding words
To ease the insertion process, I'm tempted to write the same datoms multiple times (i.e. not check if a record already exists in the database). But I'm afraid about the performance impact.
What happens in Datomic when the same datoms are added multiple times ?
Is it worth checking that a datom has already been added prior to the transaction ?
Is there a way to prevent Datomic from overriding previous datoms (i.e if a record already exists, skip the transaction) ?
Thank you for your help

What happens in Datomic when the same datoms are added multiple times ?
Is it worth checking that a datom has already been added prior to the transaction ?
Logically, a Datomic database is a sorted set of datoms, so adding the same datom several times is idempotent. However, when you're asserting a datom with a tempid, you may create a new datom for representing the same information as an old datom. This is where :db/unique comes in.
To ensure an entity does not get stored several times, you want to set the :db/unique attribute property to :db.unique/identity for the right attributes. For instance, if your schema consists of 3 attributes :word/text, :sentence/text, and :sentence/words, then :word/text and :sentence/text should be :db.unique/identity, which yields the following schema installation transaction:
[{:db/cardinality :db.cardinality/one,
:db/fulltext true,
:db/index true,
:db.install/_attribute :db.part/db,
:db/id #db/id[:db.part/db -1000777],
:db/ident :sentence/text,
:db/valueType :db.type/string,
:db/unique :db.unique/identity}
{:db/cardinality :db.cardinality/one,
:db/fulltext true,
:db/index true,
:db.install/_attribute :db.part/db,
:db/id #db/id[:db.part/db -1000778],
:db/ident :word/text,
:db/valueType :db.type/string,
:db/unique :db.unique/identity}
{:db/cardinality :db.cardinality/many,
:db/fulltext true,
:db/index true,
:db.install/_attribute :db.part/db,
:db/id #db/id[:db.part/db -1000779],
:db/ident :sentence/words,
:db/valueType :db.type/ref}]
Then the transaction for inserting inserting looks like:
[{:sentence/text "Hello World!"
:sentence/words [{:word/text "hello"
:db/id (d/tempid :db.part/user)}
{:word/text "world"
:db/id (d/tempid :db.part/user)}]
:db/id (d/tempid :db.part/user)}]
Regarding performance:
You may not need to optimize at all, but in my view, the potential performance bottlenecks of your import process are:
time spent building the transaction in the Transactor (which includes index lookups for unique attributes etc.)
time spent building the indexes.
To improve 2.: When the data you insert is sorted, indexing is faster, so an would be to insert words and sentences sorted. You can use Unix tools to sort large file even if they don't fit in memory. So the process would be:
sort sentences, insert them (:sentence/text)
extract words, sort them, insert them (:word/text)
insert word-sentence relationship (:sentence/words)
To improve 1.: indeed, it could put less pressure on the transactor to use entity ids for words that are already stored instead of the whole word text (which requires an index lookup to ensure uniqueness). One idea could be to perform that lookup on the Peer, either by leveraging parallelism and/or only for frequent words (For instance, you could insert the words from the 1st 1000 sentences, then retrieve their entity ids and keep them in a hash map).
Personally, I would not go through these optimizations until experience has shown they're necessary.

You are not at the point you need to worry about pre-optimization like this. Retail computer stores sell hard disks for about $0.05/GB, so you are talking about 50 cents worth of storage here. With Datomic's built-in storage compression, this will be even smaller. Indexes & other overhead will increase the total a bit, but it's still too small to worry about.
As with any problem, it is best to build up a solution incrementally. So, maybe do an experiment with the first 1% of your data and time the simplest possible algorithm. If that's pretty fast, try 10%. You now have a pretty good estimate of how long the whole problem will take to load the data. I'm betting that querying the data will be even faster than loading.
If you run into a roadblock after the first 1% or 10%, then you can think about reworking the design. Since you have built something concrete, you have been forced to think about the problem & solution in more detail. This is much better than hand-waving arguments & whiteboard designing. You now know a lot more about your data and possible solution implementations.
If it turns out the simplest solution won't work at a larger scale, the 2nd solution will be much easier to design & implement having had the experience of the first solution. Very rarely does the final solution spring full-formed from your mind. It is very important for any significant problem to plan on repeated refinement of the solution.
One of my favorite chapters from the seminal book
The Mythical Man Month by Fred Brooks is entitled, "Plan To Throw One Away".

What happens in Datomic when the same datoms are added multiple times?
If you are adding the word/sentence with a unique identity (:db.unique/identity) then Datomic will keep only one copy of it in the storage (i.e. single entity)
Is it worth checking that a datom has already been added prior to the transaction?
Is there a way to prevent Datomic from overriding previous datoms (i.e if a record already exists, skip the transaction)?*
Again, use :db.unique/identity, then you don't need to query for the entity id to check its existence.
For more information, please refer to here

Related

Map-Reduce with a wait

The concept of map-reduce is very familiar. It seems like a great fit for a problem I'm trying to solve, but it's either missing something (or I lack enough understanding of the concept).
I have a stream of items, structured as follows:
{
"jobId": 777,
"numberOfParts": 5,
"data": "some data..."
}
I want to do a map-reduce on many such items.
My mapping operation is straightforward - take the jobId.
My reduce operation is irrelevant for this phase, but all we know is that it takes multiple strings (the "some data..." part) and somehow reduces them to a single object.
The only problem is - I need all five parts of this job to complete before I can reduce all the strings into a single object. Every item has a "numberOfParts" property which indicates the number of items I must have before I apply the reduce operation. The items are not ordered, therefore I don't have a "partId" field.
Long story short - I need to apply some kind of a waiting mechanism that waits for all parts of the job to complete before initiating the reduce operation, and I need this waiting mechanism to rely on a value that exists within the payload (therefore solutions like kafka wouldn't work).
Is there a way to do that, hopefully using a single tool/framework?
I only want to write the map/reduce part and the "waiting" logic, the rest I believe should come out of the box.
**** EDIT ****
I'm currently in the design phase of the project and therefore not using any framework (such as spark, hadoop, etc...)
I asked this because I wanted to find out the best way to tackle this problem.
"Waiting" is not the correct approach.
Assuming your jobId is the key, and data contains some number of parts (zero or more), then you must have multiple reducers. One that gathers all parts of the same job, then another that processes all jobs with a collection of parts greater than or equal to numberOfParts while ignoring others

Slow insertion using Neptune and Gremlin

I'm having problems with the insertion using gremlin to Neptune.
I am trying to insert many nodes and edges, potentially hundred thousands of nodes and edges, with checking for existence.
Currently, we are using inject to insert the nodes, and the problem is that it is slow.
After running the explain command, we figured out that the problem was the coalesce and the where steps - it takes more than 99.9% of the run duration.
I want to insert each node and edge only if it doesn’t exist, and that’s why I am using the coalesce and where steps.
For example, the query we use to insert nodes with inject:
properties_list = [{‘uid’:’1642’},{‘uid’:’1322’}…]
g.inject(properties_list).unfold().as_('node')
.sideEffect(__.V().where(P.eq('node')).by(‘uid).fold()
.coalesce(__.unfold(), __.addV(label).property(Cardinality.single,'uid','1')))
With 1000 nodes in the graph and properties_list with 100 elements, running the query above takes around 30 seconds, and it gets slower as the number of nodes in the graph increases.
Running a naive injection with the same environment as the query above, without coalesce and where, takes less than 1 second.
I’d like to hear your suggestions and to know what are the best practices for inserting many nodes and edges (with checking for existence).
Thank you very much.
If you have a set of IDs that you want to check for existence, you can speed up the query significantly by also providing just a list of IDs to the query and calculating the intersection of the ones that exist upfront. Then, having calculated the set that need updates you can just apply them in one go. This will make a big difference. The reason you are running into problems is that the mid traversal V has a lot of work to do. In general it would be better to use actual IDs rather than properties (UID in your case). If that is not an option the same technique will work for property based IDs. The steps are:
Using inject or sideEffect insert the IDs to be found as one list and the corresponding map containing the changes to conditionally be applied in a separate map.
Find the intersection of the ones that exist and those that do not.
Using that set of non existing ones, apply the updates using the values in the set to index into your map.
Here is a concrete example. I used the graph-notebook for this but you can do the same thing in code:
Given:
ids = "['1','2','9998','9999']"
and
data = "[['id':'1','value':'XYZ'],['id':'9998','value':'ABC'],['id':'9999','value':'DEF']]"
we can do something like this:
g.V().hasId(${ids}).id().fold().as('exist').
constant(${data}).
unfold().as('d').
where(without('exist')).by('id').by()
which correctly finds the ones that do not already exist:
{'id': 9998, 'value': 'ABC'}
{'id': 9999, 'value': 'DEF'}
You can use this pattern to construct your conditional inserts a lot more efficiently (I hope :-) ). So to add the new vertices you might do:
g.V().hasId(${ids}).id().fold().as('exist').
constant(${data}).
unfold().as('d').
where(without('exist')).by('id').by().
addV('test').
property(id,select('d').select('id')).
property('value',select('d').select('value'))
v[9998]
v[9999]
As a side note, we are adding two new steps to Gremlin - mergeV and mergeE that will allow this to be done much more easily and in a more declarative style. Those new steps should be part of the TinkerPop 3.6 release.

How would I merge related records in apache beam / dataflow, based on hundreds of rules?

I have data I have to join at the record level. For example data about users is coming in from different source systems but there is not a common primary key or user identifier
Example Data
Source System 1:
{userid = 123, first_name="John", last_name="Smith", many other columns...}
Source System 2:
{userid = EFCBA-09DA0, fname="J.", lname="Smith", many other columns...}
There are about 100 rules I can use to compare one record to another
to see if customer in source system 1 is the same as source system 2.
Some rules may be able to infer record values and add data to a master record about a customer.
Because some rules may infer/add data to any particular record, the rules must be re-applied again when a record changes.
We have millions of records per day we'd have to unify
Apache Beam / Dataflow implementation
Apache beam DAG is by definition acyclic but I could just republish the data through pubsub to the same DAG to make it a cyclic algorithm.
I could create a PCollection of hashmaps that continuously do a self join against all other elements but this seems it's probably an inefficient method
Immutability of a PCollection is a problem if I want to be constantly modifying things as it goes through the rules. This sounds like it would be more efficient with Flink Gelly or Spark GraphX
Is there any way you may know in dataflow to process such a problem efficiently?
Other thoughts
Prolog: I tried running on subset of this data with a subset of the rules but swi-prolog did not seem scalable, and I could not figure out how I would continuously emit the results to other processes.
JDrools/Jess/Rete: Forward chaining would be perfect for the inference and efficient partial application, but this algorithm is more about applying many many rules to individual records, rather than inferring record information from possibly related records.
Graph database: Something like neo4j or datomic would be nice since joins are at the record level rather than row/column scans, but I don't know if it's possible in beam to do something similar
BigQuery or Spanner: Brute forcing these rules in SQL and doing full table scans per record is really slow. It would be much preferred to keep the graph of all records in memory and compute in-memory. We could also try to concat all columns and run multiple compare and update across all columns
Or maybe there's a more standard way to solving these class of problems.
It is hard to say what solution works best for you from what I can read so far. I would try to split the problem further and try to tackle different aspects separately.
From what I understand, the goal is to combine together the matching records that represent the same thing in different sources:
records come from a number of sources:
it is logically the same data but formatted differently;
there are rules to tell if the records represent the same entity:
collection of rules is static;
So, the logic probably roughly goes like:
read a record;
try to find existing matching records;
if matching record found:
update it with new data;
otherwise save the record for future matching;
repeat;
To me this looks very high level and there's probably no single 'correct' solution at this level of detail.
I would probably try to approach this by first understanding it in more detail (maybe you already do), few thoughts:
what are the properties of the data?
are there patterns? E.g. when one system publishes something, do you expect something else from other systems?
what are the requirements in general?
latency, consistency, availability, etc;
how data is read from the sources?
can all the systems publish the records in batches in files, submit them into PubSub, does your solution need to poll them, etc?
can the data be read in parallel or is it a single stream?
then the main question of how can you efficiently match a record in general will probably look different under different assumptions and requirements as well. For example I would think about:
can you fit all data in memory;
are your rules dynamic. Do they change at all, what happens when they do;
can you split the data into categories that can be stored separately and matched efficiently, e.g. if you know you can try to match some things by id field, some other things by hash of something, etc;
do you need to match against all of historical/existing data?
can you have some quick elimination logic to not do expensive checks?
what is the output of the solution? What are the requirements for the output?

Storm and stop words

I am new in storm framework(https://storm.incubator.apache.org/about/integrates.html),
I test locally with my code and I think If I remove stop words, it will perform well, but i search on line and I can't see any example that removing stopwords in storm.
If the size of the stop words list is small enough to fit in memory, the most straighforward approach would be to simply filter the tuples with an implementation of storm Filter that knows that list. This Filter could possibly poll the DB every so often to get the latest list of stop words if this list evolves over time.
If the size of the stop words list is bigger, then you can use a QueryFunction, called from your topology with the stateQuery function, which would:
receive a batch of tuples to check (say 10000 at a time)
build a single query from their content and look up corresponding stop words in persistence
attach a boolean to each tuple specifying what to with each one
+ add a Filter right after that to filter based on that boolean.
And if you feel adventurous:
Another and faster approach would be to use a bloom filter approximation. I heard that Algebird is meant to provide this kind of functionality and targets both Scalding and Storm (how cool is that?), but I don't know how stable it is nor do I have any experience in practically plugging it into Storm (maybe Sunday if it's rainy...).
Also, Cascading (which is not directly related to Storm but has a very similar set of primitive abstractions on top of map reduce) suggests in this tutorial a method based on left joins. Such joins exist in Storm and the right branch could possibly be fed with a FixedBatchSpout emitting all stop words every time, or even a custom spout that reads the latest version of the list of stop words from persistence every time, so maybe that would work too? Maybe? This also assumes the size of the stop words list is relatively small though.

JPA 2.0: Batching queries with IN clause

I am looking for a strategy to batch all my queries (with IN clause) to overcome the restrictions by databases on IN clause (See here).
I usually get list of size 100000 to 305000. So, this has become very important to tackle.
I have tried two strategies so far.
Strategy 1:
Create an entity and hence a table with one column to hold such values (can we create temp tables on the fly with JPA 2.0 vendor-independent?) and use the data from the temp table as a subquery to the original query before eventually cleaning up the temp table.
Advantage: Very performant queries. Really quick, I must admit for the numbers I have mentioned, it was mostly under a minute.
Possible drawback: Use of temp table which is actually a permanent one in my case thus far.
Strategy 2:
Calculate the batch size for the given input list and for each batch execute the query and accumulate the result.
Advantage: No temp tables. Easy for any threads within the same transaction.
Disadvantage: A big disadvantage is amount of time it takes to execute all the batches. For the mentioned numbers, this is at an unacceptable level at the moment. Takes anything between 5 to 15 mins!
I would appreciate any feedback, suggestions or improvements from all you JPA gurus.
Thanks.
I only tested up to 50,000 integers but I have some pretty good performance data around splitting large lists using various methods, with CLR and a numbers table leading the pack at the higher end:
Splitting a list of integers : another roundup
Not sure if you are using integers or strings but the results should be roughly equivalent.
As an aside, I'll confess I have no idea what JPA 2.0 is, but I assume you can control the format of the lists that it sends to SQL Server.