I'm trying to find the best data model to adapt a very big mysql table in Cassandra.
This table is structured like this:
CREATE TABLE big_table (
social_id,
remote_id,
timestamp,
visibility,
type,
title,
description,
other_field,
other_field,
...
)
A page (which is not here) can contain many socials, which can contain many remote_ids.
Social_id is the partitioning key, remote_id and timestamp are the clustering key: "Remote_id" gives unicity, "Time" is used to order the results. So far so good.
The problem is that users can also search on their page contents, filtering by one or more socials, one or more types, visibility (could be 0,1,2), a range of dates or even nothing at all.
Plus, based on the filters, users should be able to set visibility.
I tried to handle this case, but I really can find a sustainable solution.
The best I've got is to create another table, which I need to keep up with the original one.
This table will have:
page_id: partition key
timestamp, social_id, type, remote_id: clustering key
Plus, create a Materialized View for each combination of filters, which is madness.
Can I avoid creating the second table? What wuold be the best Cassandra model in this case? Should I consider switching to other technologies?
I start from last questions.
> What would be the best Cassandra model in this case?
As stated in Cassandra: The Definitive Guide, 2nd edition (which I highly recommend to read before choosing or using Cassandra),
In Cassandra you don’t start with the data model; you start with the query model.
You may want to read an available chapter about data design at Safaribooksonline.com. Basically, Cassandra wants you to think about queries only and don't care about normalization.
So the answer on
> Can I avoid creating the second table?
is You shouldn't avoiding it.
> Should I consider switching to other technologies?
That depends on what you need in terms of replication and partitioning. You may end up creating master-master synchronization based on RDBMS or something else. In Cassandra, you'll end up with duplicated data between tables and that's perfectly normal for it. You trade disk space in exchange for reading/writing speed.
> how to filter and update a big table dynamically?
If after all of the above you still want to use normalized data model in Cassandra, I suggest you look on secondary indexes at first and then move on to custom indexes like Lucene index.
I'm trying to decide which way to go if I have an app that needs to be able to change the db schema based on the user input.
For example, if I have a "car" object that contains car properties, like year, model, # of doors etc, how do I store it in the DB in such a way, that the user should be able to add new properties?
I read about EAV tables and they seem right for this thing, but the problem is that queries will get pretty complicated when I try to get a list of cars filtered by a set of properties.
Could I generate the tables dynamically instead? I see that Sqlite has support for ADD COLUMN, but how fast is it when the table reaches many records? And it looks like there's no way to remove a column. I have to create a new table without the column I want to remove, and copy the data from the old table. That's certainly slow on large tables :(
I will assume that SQLite (or another relational DBMS) is a requirement.
EAVs
I have worked with EAVs and generic data models, and I can say that the data model is very messy and hard to work with in the long run.
Lets say that you design a datamodel with three tables: entities, attributes, and _entities_attributes_:
CREATE TABLE entities
(entity_id INTEGER PRIMARY KEY, name TEXT);
CREATE TABLE attributes
(attribute_id INTEGER PRIMARY KEY, name TEXT, type TEXT);
CREATE TABLE entity_attributes
(entity_id INTEGER, attribute_id INTEGER, value TEXT,
PRIMARY KEY(entity_id, attribute_id));
In this model, the entities table will hold your cars, the attributes table will hold the attributes that you can associate to your cars (brand, model, color, ...) and its type (text, number, date, ...), and the _entity_attributes_ will hold the values of the attributes for a given entity (for example "red").
Take into account that with this model you can store as many entities as you want and they can be cars, houses, computers, dogs or whatever (ok, maybe you need a new field on entities, but it's enough for the example).
INSERTs are pretty straightforward. You only need to insert a new object, a bunch of attributes and its relations. For example, to insert a new entity with 3 attributes you will need to execute 7 inserts (one for the entity, three more for the attributes, and three more for the relations.
When you want to perform an UPDATE, you will need to know what is the entity that you want to update, and update the desired attribute joining with the relation between the entity and its attributes.
When you want to perform a DELETE, you will also need to need to know what is the entity you want to delete, delete its attributes, delete the relation between your entity and its attributes and then delete the entity.
But when you want to perform a SELECT the thing becomes nasty (you need to write really difficult queries) and the performance drops horribly.
Imagine a data model to store car entities and its properties as in your example (say that we want to store brand and model). A SELECT to query all your records will be
SELECT brand, model FROM cars;
If you design a generic data model as in the example, the SELECT to query all your stored cars will be really difficult to write and will involve a 3 table join. The query will perform really bad.
Also, think about the definition of your attributes. All your attributes are stored as TEXT, and this can be a problem. What if somebody makes a mistake and stores "red" as a price?
Indexes are another thing that you could not benefit of (or at least not as much as it would be desirable), and they are very neccesary as the data stored grows.
As you say, the main concern as a developer is that the queries are really hard to write, hard to test and hard to maintain (how much would a client have to pay to buy all red, 1980, Pontiac Firebirds that you have?), and will perform very poorly when the data volume increases.
The only advantage of using EAVs is that you can store virtually everything with the same model, but is like having a box full of stuff where you want to find one concrete, small item.
Also, to use an argument from authority, I will say that Tom Kyte argues strongly against generic data models:
http://tkyte.blogspot.com.es/2009/01/this-should-be-fun-to-watch.html
https://asktom.oracle.com/pls/asktom/f?p=100:11:0::::P11_QUESTION_ID:10678084117056
Dynamic columns in database tables
On the other hand, you can, as you say, generate the tables dynamically, adding (and removing) columns when needed. In this case, you can, for example create a car table with the basic attributes that you know that you will use and then add columns dynamically when you need them (for example the number of exhausts).
The disadvantage is that you will need to add columns to an existing table and (maybe) build new indexes.
This model, as you say, also has another problem when working with SQLite as there's no direct way to delete columns and you will need to do this as stated on http://www.sqlite.org/faq.html#q11
BEGIN TRANSACTION;
CREATE TEMPORARY TABLE t1_backup(a,b);
INSERT INTO t1_backup SELECT a,b FROM t1;
DROP TABLE t1;
CREATE TABLE t1(a,b);
INSERT INTO t1 SELECT a,b FROM t1_backup;
DROP TABLE t1_backup;
COMMIT;
Anyway, I don't really think that you will need to delete columns (or at least it will be a very rare scenario). Maybe someone adds the number of doors as a column, and stores a car with this property. You will need to ensure that any of your cars have this property to prevent from losing data before deleting the column. But this, of course depends on your concrete scenario.
Another drawback of this solution is that you will need a table for each entity you want to store (one table to store cars, another to store houses, and so on...).
Another option (pseudo-generic model)
A third option could be to have a pseudo-generic model, with a table having columns to store id, name, and type of the entity, and a given (enough) number of generic columns to store the attributes of your entities.
Lets say that you create a table like this:
CREATE TABLE entities
(entity_id INTEGER PRIMARY KEY,
name TEXT,
type TEXT,
attribute1 TEXT,
attribute1 TEXT,
...
attributeN TEXT
);
In this table you can store any entity (cars, houses, dogs) because you have a type field and you can store as many attributes for each entity as you want (N in this case).
If you need to know what the attribute37 stands for when type is "red", you would need to add another table that relates the types and attributes with the description of the attributes.
And what if you find that one of your entities needs more attributes? Then simply add new columns to the entities table (attributeN+1, ...).
In this case, the attributes are always stored as TEXT (as in EAVs) with it's disadvantages.
But you can use indexes, the queries are really simple, the model is generic enough for your case, and in general, I think that the benefits of this model are greater than the drawbacks.
Hope it helps.
Follow up from the comments:
With the pseudo-generic model your entities table will have a lot of columns. From the documentation (https://www.sqlite.org/limits.html), the default setting for SQLITE_MAX_COLUMN is 2000. I have worked with SQLite tables with over 100 columns with great performance, so 40 columns shouldn't be a big deal for SQLite.
As you say, most of your columns will be empty for most of your records, and you will need to index all of your colums for performance, but you can use partial indexes (https://www.sqlite.org/partialindex.html). This way, your indexes will be small, even with a high number of rows, and the selectivity of each index will be great.
If you implement a EAV with only two tables, the number of joins between tables will be less than in my example, but the queries will still be hard to write and maintain, and you will need to do several (outer) joins to extract data, which will reduce performance, even with a great index, when you store a lot of data. For example, imagine that you want to get the brand, model and color of your cars. Your SELECT would look like this:
SELECT e.name, a1.value brand, a2.value model, a3.value color
FROM entities e
LEFT JOIN entity_attributes a1 ON (e.entity_id = a1.entity_id and a1.attribute_id = 'brand')
LEFT JOIN entity_attributes a2 ON (e.entity_id = a2.entity_id and a2.attribute_id = 'model')
LEFT JOIN entity_attributes a3 ON (e.entity_id = a3.entity_id and a3.attribute_id = 'color');
As you see, you would need one (left) outer join for each attribute you want to query (or filter). With the pseudo-generic model the query will be like this:
SELECT name, attribute1 brand, attribute7 model, attribute35 color
FROM entities;
Also, take into account the potential size of your _entity_attributes_ table. If you can potentially have 40 attributes for each entity, lets say that you have 20 not null for each of them. If you have 10,000 entities, your _entity_attributes_ table will have 200,000 rows, and you will be querying it using one huge index. With the pseudo-generic model you will have 10,000 rows and one small index for each column.
It all depends on the way in which your application needs to reason about the data.
If you need to run queries which need to do complicated comparisons or joins on data whose schema you don't know in advance, SQL and the relational model are rarely a good fit.
For instance, if your users can set up arbitrary data entities (like "car" in your example), and then want to find cars whose engine capacity is greater than 2000cc, with at least 3 doors, made after 2010, whose current owner is part of the "little old ladies" table, I'm not aware of an elegant way of doing this in SQL.
However, you could achieve something like this using XML, XPath etc.
If your application has a set on data entities with known attributes, but users can extend those attributes (a common requirement for products like bug trackers), "add column" is a good solution. However, you may need to invent a custom query language to allow users to query those columns. For instance, Atlassian Jira's bug tracking solution has JQL, a SQL-like language for querying bugs.
EAV is great if your task is to store and then show data. However, even moderately complex queries become very hard in an EAV schema - imagine how you'd execute my made up example above.
For your use case, a document oriented database like MongoDB would do great.
Another option that I haven't seen mentioned above is to use denormalized tables for the extended attributes. This is a combination of the pseudo-generic model and the dynamic columns in database tables. Instead of adding columns to existing tables, you add columns or groups of columns into new tables with FK indexes to the source table. Of course, you'll want a good naming convention (car, car_attributes_door, car_attributes_littleOldLadies)
Your selection problem becomes that of applying a LEFT OUTER JOIN to include the extended attributes that you want to include.
Slower than normalized, but not as slow as EAV.
Adding new extended attributes becomes a problem of adding a new table.
Harder than EAV, easier/faster than modifying table schema.
Deleting attributes becomes a problem of dropping whole tables.
Easier/faster than modifying table schema.
These new attributes can be strongly typed.
As good as modifying table schema, faster than EAV or generic columns.
The biggest advantage to this approach that I can see is that deleting unused attributes is quite easy compared to any of the others via a single DROP TABLE command. You also have the option to later normalize often-used attributes into larger groups or into the main table using a single ALTER TABLE process rather than one for each new column you were adding as you added them, which helps with the slow LEFT OUTER JOIN queries.
The biggest disadvantage is that you're cluttering up your table list, which admittedly is often not a trivial concern. That and I'm not sure how much better LEFT OUTER JOIN's actually perform than EAV table joins. It's definitely closer to EAV join performance than normalized table performance.
If you're doing a lot of comparisons/filters of values that benefit greatly from strongly typed columns, but you add/remove these columns frequently enough to make modifying a huge normalized table intractable, this seems like a good compromise.
I would try EAV.
Adding columns based on user input doesn't sounds nice to me and you can quickly run out of capacity. Queries on very flat table can also be a problem. Do you want to create hundreds of indexes?
Instead of writing every thing to one table, I would store as many as possible common properties (price, name , color, ...) in the main table and those less common properties in an "extra" attributes table. You can always balance them later with a little effort.
EAV can performance well for small to middle sized data set. Since you want to use SQLlite, I guess it's not be a problem.
You may also want to avoid "over" normalizing your data. With the cheap storage
we currently have, you can use one table to store all "Extra" attributes, instead of two:
ent_id, ent_name, ...
ent_id, attr_name, attr_type, attr_value ...
People against EAV will say its performance is poor on large database. It's sure that it won't performance as well as normalized structure but you don't want to change structure on a 3TB table either.
I have a low quality answer, but possible, that came from HTML tags that are like : <tag width="10px" height="10px" ... />
In this dirty way you will have just one column as a varchar(max) for all properties say it Props column and you will store data in it like this:
Props
------------------------------------------------------------
Model:Model of car1|Year:2010|# of doors:4
Model:Model of car2|NewProp1:NewValue1|NewProp2:NewValue2
In this way all works will go to the programming code in business layer with using some functions like concatCustom that get an array and return a string and unconcatCustom that get a string and return an array.
For more validity of special characters like ':' and '|', I suggest '#:#' and '#|#' or something more rare for splitter part.
In a similar way you can use a text or binary field and store an XML data in the column.
Please, read first my previous question: T-SQL finding of exactly same values in referenced table
The main purpose of this question is to find out if this approach of storing of data is effective.
Maybe it would be better to get rid of PropertyValues table. And use additional PropertyValues nvarchar(max) column in Entities table instead of it. For example instead of
EntityId PropertyId PropertyValue
1 4 Val4
1 5 Val5
1 6 Val6
table, I could store such data in PropertyValues column: "4:Val4;5:Val5;6Val6"
As an alternative, I could store XML in PropertyValues column....
What do you think about the best approach here?
[ADDED]
Please, keep in mind:
Set of properties must be customizable
Objects will have dozens of properties (approximately from 20 to 120). Database will contain thousands of objects
[ADDED]
Data in PropertyValues table will be changed very often. Actually, I store configured products. For example, admin configures that clothes have attributes "type", "size", "color", "buttons type", "label type", "label location" etc... User will select values for these attributes from the system. So, PropertyValues data cannot be effectively cached.
You will hate yourself later if you implement a solution using multi-value attributes (i.e. 4:Val4;5:Val5;6Val6).
XML is marginally better because there are XQuery functions to help you pull out and parse the values. But the XML type is implemented as a CLR type in SQL Server and it can get extremely slow to work with.
The best solution to this problem is one like you have. Use the sql_variant type for the column if it could be any number of data types. Ideally you'd refactor this into multiple tables / entities so that the data type can be something more concrete.
I work with the similar project (web-shop generator). So every product has attribute and every attribute has set of values. It is different tables. And for all of this there are translations in several languages. (So exists additional tables for attributes and values translations).
Why we choose such solution? Because for every client there should be database with the same scheme. So such database scheme is very elastic.
So what about this solution. As always, "it depends" -))
Storage. If your value will be used often for different products, e.g. clothes where attribute "size" and values of sizes will be repeated often, your attribute/values tables will be smaller. Meanwhile, if values will be rather unique that repeatable (e.g. values for attribute "page count" for books), you will get a big enough table with values, where every value will be linked to one product.
Speed. This scheme is not weakest part of project, because here data will be changed rarely. And remember that you always can denormalize database scheme to prepare DW-like solution. You can use caching if database part will be slow too.
Elasticity This is the strongest part of solution. You can easily add/remove attributes and values and ever to move values from one attribute to another!
So answer on your question is not simple. If you prepare elastic scheme with unknown attributes and values, you should use different tables. I suggest to you remember about storing values in CSV strings. It is better to store it as XML (typed and indexed).
UPDATE
I think that PropertyValues will not change often , if comparing with user orders. But if you doubt, you should use denormalization tables or indexed views to speed up.Anyway, changing XML/CSV on large quantity of rows will have poor performance, so "separate table" solution looks good.
The SQL Customer Advisory Team (CAT) has a whitepaper written just for you: Best Practices for Semantic Data Modeling for Performance and Scalability. It goes through the common pitfalls of EAV modeling and recommends how to design a scalable EAV solution.
I am using SQL Server 2005 Express and Visual Studio 2008.
I have a database which has a table with 400 Columns. Things were (just about manageable) until I had to perform bi-directional sync between several databases.
I am wondering what arguments are for and against using 400 column database or 40 table database are?
The table in not normalised and comprises of mainly nvarchar(64) columns and some TEXT columns. (there are no datatypes as it was converted from text files).
There is one other table that links to this table and is a 1-1 relationship (i.e one entry relates to one entry in the 400 column table).
The table is a list files that contained parameters that are "plugged" into a application.
I look forward to your replies.
Thank you
Based on your process description I would start with something like this. The model is simplified, does not capture history, etc -- but, it is a good starting point. Note: parameter = property.
- Setup is a collection of properties. One setup can have many properties, one property belongs to one setup only.
- Machine can have many setups, one setup belongs to one machine only.
- Property is of a specific type (temperature, run time, spindle speed), there can be many properties of a certain type.
- Measurement and trait are types of properties. Measurement is a numeric property, like speed. Trait is a descriptive property, like color or some text.
For having a wide table:
Quick to report on as it's presumably denormalized and so no joins are needed.
Easy to understand for end-consumers as they don't need to hold a data model in their heads.
Against having a wide table:
Probably need to have multiple composite indexes to get good query performance
More difficult to maintain data consistency i.e. need to update multiple rows when data changes if that data is on multiple rows
As you're having to update multiple rows and maintain multiple indexes, concurrent performance for updates may become an issue as locks escalate.
You might end up with records with loads of nulls in columns if the attribute isn't relevant to the entity on that row which can make handling results awkward.
If lazy developers do a SELECT * from the table you end up dragging loads of data across the network, so you generally have to maintain suitable subset views.
So it all really depends on what you're doing. If the main purpose of the table is OLAP reporting and updates are infrequent and affect few rows then perhaps a wide, denormalized table is the right thing to have. In an OLTP environment then it's probably not and you should prefer narrower tables. (I generally design in 3NF and then denormalize for query performance as I go along.)
You could always take the approach of normalizing and providing a wide-view for readers if that's what they want to see.
Without knowing more about the situation it's not really possible to say more about the pros and cons in your particular circumstance.
Edit:
Given what you've said in your comments, have you considered just having a long & skinny name=value pair table so you'd just have UserId, PropertyName, PropertyValue columns? You might want to add in some other meta-attributes into it too; timestamp, version, or whatever. SQL Server is quite efficient at handling these sorts of tables so don't discount a simple solution like this out-of-hand.
I want to store certain items in the database with variable amount of properties.
For example:
An item can have 'url' and 'pdf' property both others do not en instead have 'image' and 'location' properties.
So the problem is an some items can have some properties and others a lot.
How would you design this database. How to make it searchable and performant?
What would the schema look like?
Thanks!
What you are after has a name - Entity Attribute Value (EAV). It is "a data model that is used in circumstances where the number of attributes (properties, parameters) that can be used to describe a thing (an "entity" or "object") is potentially very vast, but the number that will actually apply to a given entity is relatively modest."
If you are not necessarily tied to SQL, a triple store is designed for precisely this task. Most are designed to be queried with the SPARQL query language.
That sounds like a perfect job for a document database.
Start with your object (item) and create a table for items. Your item can have 1 or many attributes or none at all right? So set up a table of attributes with unique ids. Now set up a table that holds many items (some can duplicate) and many attributes (can duplicate as well)
Item
ItemID
ItemDescription
...
Attributes
AttributeID
AttributeDescription
...
ItemAttributes
rowID
ItemID
AttributeID
Now when you want to query you can simply join the tables and filter however you desire...
The Entity Attribute Value (EAV) model is very flexible. The semantic web and its query language sparql are based on EAV too. But some people don't like it because there is a performance penalty with this model.
Start with doing some high load performance tests on your database. Don't do them when you are done coding, because then it is too late.
edit: Focus on the speed of you select statements. Users expect quick results when they search.
I have designed tables like this in the past to have the following fields:
id
type
subtype
value
And then I would have another table that would define the type and subtypes used, and possibly give the datatype for that type and subtype combination so that you could programatically enforce it.
Its not pretty, and you don't want to do it unless you have to. But its the best way I have found when you do.
update: even if you leave subtype blank, I find its a good thing to have, because its too often that you want to subcategorize something that already exists. Example you create type: address, now you need mailing address and billing address and physical address.
For this kind of scenario's I use the XML-type column in MS SQL 2005...
you'll have all the advantages of XML + SQL. That is use an XPath expression as part of an SQL-statement.
It's a feature of MS SQL 2005, I am not sure which other RDBMS support this.
I am not sure what the implications are performance wise.
Create a properties table with the following fields:
item_id int(or whatever the ID type is in the item table)
property_name varchar(500)
property_value varchar(500)
Set a foreign key between item_id and the item's id field, and you're done.
That's how you do a many-to-one relationship in SQL.
Looks like an "items" table with primary key "item_id", a "properties" table with primary key "property_id" and a foreign key "item_id" with the "items" table. "properties" will have columns "name" and "value", both of type varchar.
Performant? Don't know.