Apologies if this has been covered thoroughly in the past - I've seen some related posts but haven't found anything that satisfies me with regards to this specific scenario.
I've been recently looking over a relatively simple game with around 10k players. In the game you can catch and breed pets that have certain attributes (i.e. wings, horns, manes). There's currently a table in the database that looks something like this:
-------------------------------------------------------------------------------
| pet_id | wings1 | wings1_hex | wings2 | wings2_hex | horns1 | horns1_hex | ...
-------------------------------------------------------------------------------
| 1 | 1 | ffffff | NULL | NULL | 2 | 000000 | ...
| 2 | NULL | NULL | NULL | NULL | NULL | NULL | ...
| 3 | 2 | ff0000 | 1 | ffffff | 3 | 00ff00 | ...
| 4 | NULL | NULL | NULL | NULL | 1 | 0000ff | ...
etc...
The table goes on like that and currently has 100+ columns, but in general a single pet will only have around 1-8 of these attributes. A new attribute is added every 1-2 months which requires table columns to be added. The table is rarely updated and read frequently.
I've been proposing that we move to a more vertical design scheme for better flexibility as we want to start adding larger volumes of attributes in the future, i.e.:
----------------------------------------------------------------
| pet_id | attribute_id | attribute_color | attribute_position |
----------------------------------------------------------------
| 1 | 1 | ffffff | 1 |
| 1 | 3 | 000000 | 2 |
| 3 | 2 | ffffff | 1 |
| 3 | 1 | ff0000 | 2 |
| 3 | 3 | 00ff00 | 3 |
| 4 | 3 | 0000ff | 1 |
etc...
The old developer has raised concerns that this will create performance issues as users very frequently search for pets with specific attributes (i.e. must have these attributes, must have at least one in this colour or position, must have > 30 attributes). Currently the search is quite fast as there are no JOINS required, but introducing a vertical table would presumably mean an additional join for every attribute searched and would also triple the number of rows or so.
The first part of my question is if anyone has any recommendations with regards to this? I'm not particularly experienced with database design or optimisation.
I've run tests for a variety of cases but they've been largely inconclusive - the times vary quite significantly for all of the queries that I ran (i.e. between half a second and 20+ seconds), so I suppose the second part of my question is whether there's a more reliable way of profiling query times than using microtime(true) in PHP.
Thanks.
This is called the Entity-Attribute-Value-Model, and relational database systems are really not suited for it at all.
To quote someone who deems it one of the five errors not to make:
So what are the benefits that are touted for EAV? Well, there are none. Since EAV tables will contain any kind of data, we have to PIVOT the data to a tabular representation, with appropriate columns, in order to make it useful. In many cases, there is middleware or client-side software that does this behind the scenes, thereby providing the illusion to the user that they are dealing with well-designed data.
EAV models have a host of problems.
Firstly, the massive amount of data is, in itself, essentially unmanageable.
Secondly, there is no possible way to define the necessary constraints -- any potential check constraints will have to include extensive hard-coding for appropriate attribute names. Since a single column holds all possible values, the datatype is usually VARCHAR(n).
Thirdly, don't even think about having any useful foreign keys.
Finally, there is the complexity and awkwardness of queries. Some folks consider it a benefit to be able to jam a variety of data into a single table when necessary -- they call it "scalable". In reality, since EAV mixes up data with metadata, it is lot more difficult to manipulate data even for simple requirements.
The solution to the EAV nightmare is simple: Analyze and research the users' needs and identify the data requirements up-front. A relational database maintains the integrity and consistency of data. It is virtually impossible to make a case for designing such a database without well-defined requirements. Period.
The table goes on like that and currently has 100+ columns, but in general a single pet will only have around 1-8 of these attributes.
That looks like a case for normalization: Break the table into multiple, for example one for horns, one for wings, all connected by foreign key to the main entity table. But do make sure that every attribute still maps to one or more columns, so that you can define constraints, data types, indexes, and so on.
Do the join. The database was specifically designed to support joins for your use case. If there is any doubt, then benchmark.
EDIT: A better way to profile the queries is to run the query directly in the MySQL interpretter on the CLI. It will give you the exact time that it took to run the query. The PHP microtime() function will also introduce other latencies (Apache, PHP, server resource allocation, network if connection to a remote MySQL instance, etc).
What you are proposing is called 'normalization'. This is exactly what relational databases were made for - if you take care of your indexes, the joins will run almost as fast as if the data were in one table.
Actually, they might even go faster: instead of loading 1 table row with 100 columns, you can just load the columns you need. If a pet only has 8 attributes, you only load those 8.
This question is a very subjective. If you have the resources to update the middleware to reflect the column that has been added then, by all means, go with horizontal there is nothing safer and easier to learn than a fixed structure. One thing to remember, anytime you update a tables structure you have to update each one of its dependencies unless there is some catch-all like *, which I suggest you stay aware from unless you are just dumping data to a screen and order of columns is irrelevant.
With that said, Verticle is the way to go if you don't have all of your requirements in place or don't have the desire to update code in n number of areas. Most of the time you just need storage containers to store data. I would segregate things like numbers, dates, binary, and text in separate columns to preserve some data integrity, but there is nothing wrong with verticle storage, as long as you know how to formulate and structure queries to bring back the data in the appropriate format.
FYI, Wordpress uses verticle data storage for majority of the dynamic content it has to store for the millions of uses it has.
First thing from Database point of view is that your data should be grow vertically not in horizontal way. So, adding a new column is not a good design at all. Second thing, this is very common scenario in DB design. And the way to solve this you have to create three tables. 1st is of Pets, 2nd is of Attributes and 3rd is mapping table between theres two. Here is the example:
Table 1 (Pet)
Pet_ID | Pet_Name
1 | Dog
2 | Cat
Table 2 (Attribute)
Attribute_ID | Attribute_Name
1 | Wings
2 | Eyes
Table 3 (Pet_Attribute)
Pet_ID | Attribute_ID | Attribute_Value
1 | 1 | 0
1 | 2 | 2
About Performance:
Pet_ID and Attribute_ID are the primary keys which are indexed (http://developer.mimer.com/documentation/html_92/Mimer_SQL_Engine_DocSet/Basic_concepts4.html), so the search is very fast. And this is the right way to sovle the problem. Hope, now it will be clear to you.
Related
i´m new here and have a project where i have a performance problem that seems hard to fix. I have created a search for objects that have availibilities means a very simple structure:
ObjectID | Date | Number of available objects
---------------------------------------------
Object1 | 01.01.2019 | 1
Object1 | 02.01.2019 | 1
Object1 | 03.01.2019 | 0
Object1 | 04.01.2019 | 1
Object1 | 05.01.2019 | 1
Object2 | 01.01.2019 | 1
Object2 | 02.01.2019 | 1
Object2 | 03.01.2019 | 0
Object2 | 04.01.2019 | 1
Object2 | 05.01.2019 | 1
I´m working with mysql and php
A typical query would be:
Which objects are available between 01.01.2019 - 28.02.2019 10 days available in a row.
It´s not really hard to make it working with mysql but once you have more then 10 users using the searchfunction the server load becomes extremly high eventough the table is optimised (indexes etc.) The server has 2 cores with 4 GB of RAM.
I also tried to store the dates comma separated per object in a table and let the application search but that creates extrem high traffic between application and database which is also not a real solution.
In total we have around 20.000 Objects and availabilities stored for max. 500 days so we have around 10.000.000 datasets in my first solution.
Does anybody have and idea what´s the most efficient way toDo this ?
(How to store it to make search fast ?)
For this project i sadly can not cache the searches.
Thanks for you help and Kind Regards, Christoph
Don't store dates in 28.02.2019 format. Flip it over, then use a DATE datatype in the table. Please provide SHOW CREATE TABLE.
What is your algorithm for searching?
The header says "number of objects", yet the values seem to be only 0 or 1, as if it is a boolean flag??
What is the maximum timespan? (If under 64, there are bit-oriented tricks we could play.)
By looking at adjacent rows (cf LAG(), if using MySQL 8.0), decide when an object changes state. Save those dates.
From that, it is one more hop to get "how many consecutive days" starting at one of those dates. This will be a simple query, and very fast if you have a suitable composite index.
Take a look at the items table below, as you can see this table is not normalized. Name should in a separate table to normalize it.
mysql> select * from items;
+---------+--------+-----------+------+
| item_id | cat_id | name | cost |
+---------+--------+-----------+------+
| 1 | 102 | Mushroom | 5.00 |
| 2 | 2 | Mushroom | 5.40 |
| 3 | 173 | Pepperoni | 4.00 |
| 4 | 109 | Chips | 1.00 |
| 5 | 35 | Chips | 1.00 |
+---------+--------+-----------+------+
This table is not normalize because on the backend Admin site, staff simply select a category and type in the item name to add data quickly. It is very quick. There are hundreds of same item name but the cost is not always the same.
If I do normalize this table to something like this:
mysql> select * from items;
+---------+--------+--------------+------+
| item_id | cat_id | item_name_id | cost |
+---------+--------+--------------+------+
| 1 | 102 | 1 | 5.00 |
| 2 | 2 | 1 | 5.40 |
| 3 | 173 | 2 | 4.00 |
| 4 | 109 | 3 | 1.00 |
| 5 | 35 | 3 | 1.00 |
+---------+--------+--------------+------+
mysql> select * from item_name;
+--------------+-----------+
| item_name_id | name |
+--------------+-----------+
| 1 | Mushroom |
| 2 | Pepperoni |
| 3 | Chips |
+--------------+-----------+
Now how can I add item (data) on the admin backend (data entry point of view) because this table has been normalized? I don't want like a dropdown to select item name - there will be thousands of different item name - it will take a lot of of time to find the item name and then type in the cost.
There need to be a way to add item/data quick as possible. What is the solution to this? I have developed backend in PHP.
Also what is the solution for editing the item name? Staff might rename the item name completely for example: Fish Kebab to Chicken Kebab and that will effect all the categories without realising it. There will be some spelling mistake that may need correcting like F1sh Kebab which should be Fish Kebab (This is useful when the tables are normalized and I will see item name updated every categories).
I don't want like a dropdown to select item name - there will be thousands of different item name - it will take a lot of of time to find the item name and then type in the cost.
There are options for selecting existing items other than drop down boxes. You could use autocompletion, and only accept known values. I just want to be clear there are UI friendly ways to achieve your goals.
As for whether to do so or not, that is up to you. If the product names are varied slightly, is that a problem? Can small data integrity issues like this be corrected with batch jobs or similar if they are a problem?
Decide what your data should look like first, based on the design of your system. Worry about the best way to structure a UI after you've made that decision. Like I said, there are usable ways to design UI regardless of your data structuring.
I think you are good to go with your current design, for you name is the product name and not the category name, you probably want to avoid cases where renaming a single product would rename too many of them at once.
Normalization is a good thing but you have to measure it against your specific needs and in this case I really would not add an extra table item_name as you shown above.
just my two cents :)
What are the dependencies supposed to be represented by your table? What are the keys? Based on what you've said I don't see how your second design is any more normalized that your first.
Presumably the determinants of "name" in the first design are the same as the determinants of "item_name_id" in the second? If so then moving name to another table won't make any difference to the normal forms satisified by your items table.
User interface design has nothing to do with database design. You cannot let the UI drive the database design and expect sensible results.
You need to validate the data and check for existence prior to adding it to see if it's a new value.
$value = $_POST['userSubmittedValue']
//make sure you sanitize the variable (never trust user input)
$query = SELECT item_name_id
FROM item_name
WHERE name='$value';
$result = mysql_query($query);
$row = mysql_fetch_row($result);
if(!empty($row))
{
//add the record with the id from $row['item_name_id'] to items table
}
else
{
//this will be a new value so run queries to add the new value to both items and item_name tables
}
There need to be a way to add item/data quick as possible. What is the
solution to this? I have developed backend in PHP.
User interface issues and database structure are separate issues. For a given database structure, there are usually several user-friendly ways to present and change the data. Data integrity comes from the database. The user interface just needs to know where to find unique values. The programmer decides how to use those unique values. You might use a drop-down list, pop up a search form, use autocomplete, compare what the user types to the elements in an array, or query the database to see whether the value already exists.
From your description, it sounds like you had a very quick way to add data in the first place: "staff simply select a category and type in the item name to add data quickly". (Replacing "mushroom" with '1' doesn't have anything to do with normalization.)
Also what is the solution for editing the item name? Staff might
rename the item name completely for example: Fish Kebab to Chicken
Kebab and that will effect all the categories without realising it.
You've allowed the wrong person to edit item names. Seriously.
This kind of issue arises in every database application. Allow only someone trained and trustworthy to make these kinds of changes. (See your dbms docs for GRANT and REVOKE. Also take a look at ON UPDATE RESTRICT.)
In our production database at work, I can insert new states (for the United States), and I can change existing state names to whatever I want. But if I changed "Alabama" to "Kyrgyzstan", I'd get fired. Because I'm supposed to know better than to do stuff like that.
But even though I'm the administrator, I can't edit a San Francisco address and change its ZIP code to '71601'. The database "knows" that '71601' isn't a valid ZIP code for San Francisco. Maybe you can add a table or two to your database, too. I can't tell from your description whether something like that would help you.
On systems where I'm not the administrator, I'd expect to have no permissions to insert rows into the table of states. In other tables, I might have permission to insert rows, but not to update or delete them.
There will be some spelling mistake that may need correcting like F1sh
Kebab which should be Fish Kebab
The lesson is the same. Some people should be allowed to update items.name, and some people should not. Revoke permissions, restrict cascading updates, increase data integrity using more tables, or increase training.
What issues are associated with maintaining multiple tress within a single table?
The motivation for having multiple trees is to avoid excessive updates to all nodes when inserting a node at the start. Each of the trees are completely separate entities.
Example Table:
tree_id | id | lft | rgt | parent_id | various fields . . .
---------------------------------------------------------------------
1 | 1 | 1 | 4 | NULL | ...
1 | 2 | 2 | 3 | 1 | ...
2 | 3 | 1 | 4 | NULL | ...
2 | 4 | 2 | 3 | 3 | ...
It's very common to store multiple trees in one table, just have to make sure the values that comprise a tree are stored correctly, otherwise it'll lead to data integrity issues like nonsensical tree constructions.
Suppose we have a binary tree, (like the one in your example). If a tree was 5 depth. ((2^n)-1) = (2^5 - 1) nodes would exist or 31 rows in the database which is trivial. Even at 10 depth it's still a small amount of rows, but would be a rather ginormous tree. And so having multiple trees, X, in there would be X((2^n)-1) = rows... in the database which isn't bad. So potentially a hundred trees could exist in one table and would only be 100k rows which is relatively small.
Additionally, suppose every new tree constructed was stored in its own table, then very quickly, the database would be filled with quite a bit of tables over time to match the number of trees that exist. And it just seems like not a good idea to make extra tables that are unneeded, adds unneeded complexity in the code side to have to access these multiple tables.
Looking at your table in detail, it doesn't quite look right in terms of columns, but I'm sure that table example is just something thrown up quickly to show us what you mean.
tree_id, node_id, left_node_id, right_node_id, various_fields...
Um, be sure to index those _id fields.
This is for an upcoming project. I have two tables - first one keeps tracks of photos, and the second one keeps track of the photo's rank
Photos:
+-------+-----------+------------------+
| id | photo | current_rank |
+-------+-----------+------------------+
| 1 | apple | 5 |
| 2 | orange | 9 |
+-------+-----------+------------------+
The photo rank keeps changing on a regular basis, and this is the table that tracks it:
Ranks:
+-------+-----------+----------+-------------+
| id | photo_id | ranks | timestamp |
+-------+-----------+----------+-------------+
| 1 | 1 | 8 | * |
| 2 | 2 | 2 | * |
| 3 | 1 | 3 | * |
| 4 | 1 | 7 | * |
| 5 | 1 | 5 | * |
| 6 | 2 | 9 | * |
+-------+-----------+----------+-------------+ * = current timestamp
Every rank is tracked for reporting/analysis purpose.
[Edit] Users will have access to the statistics on demand.
I talked to someone who has experience in this field, and he told me that storing ranks like above is the way to go. But I'm not so sure yet.
The problem here is data redundancy. There are going to be tens of thousands of photos. The photo rank changes on a hourly basis (many times- within minutes) for recent photos but less frequently for older photos. At this rate the table will have millions of records within months. And since I do not have experience in working with large databases, this makes me a little nervous.
I thought of this:
Ranks:
+-------+-----------+--------------------+
| id | photo_id | ranks |
+-------+-----------+--------------------+
| 1 | 1 | 8:*,3:*,7:*,5:* |
| 2 | 2 | 2:*,9:* |
+-------+-----------+--------------------+ * = current timestamp
That means some extra code in PHP to split the rank/time (and sorting), but that looks OK to me.
Is this a correct way to optimize the table for performance? What would you recommend?
The first one. Period.
Actually you'll lose much more. A timestamp stored in the int column will occupy only 4 bytes of space.
While the same timestamp stored in the string format will take 10 bytes.
Your first design is correct for a relational database. The redundancy in the key columns is preferable because it gives you a lot more flexibility in how you validate and query the rankings. You can do sorts, counts, averages, etc. in SQL without having to write any PHP code to split your string six ways from Sunday.
It sounds like you would like to use a non-SQL database like CouchDB or MongoDB. These would allow you to store a semi-structured list of rankings right in the record for the photo, and subsequently query the rankings efficiently. With the caveat that you don't really know that the rankings are in the right format, as you do with SQL.
I would stick with your first approach. In the second you will have a lot of data stored in the row, as time goes by it gets more ranks! That is, if a photo gets thousands and thousands of rankings.
The first approach is also more maintainable, that is, if you wish to delete a rank.
I'd think the database 'hit' of over normalistion (querying the ranks table over and over) is nicely avoided by 'caching' the last rank in current_rank. It does not really matter ranks is growing tremendously if it is seldom queried (analyis / reporting you said), never updated but just gets records inserted at the end: even a very light box would have no problem having millions of rows in that table.
You alternative would require lots of updates on different locations on the disk, possibly resulting in degraded performance.
Of course, if you need all the old data, and always by photo_id, you could plan a scheduled run to another table rankings_old, possibly with photo_id, year,month, rankings (including timestamps) when a month is over, so retrieving old data stays easily possible, but there are no updates needed in rankings_old or rankings, only inserts at the end of the table.
And take it from me: millions of records in a pure logging table should be absolutely no problem.
Normalized data or not normalized data. You will find thousands of articles about that. :)
It really depends of your needs.
If you want to build your database only with performance (speed or RAM consumption or...) in mind you should only trust the numbers. To do that you have to profile your queries with the expected data "volume" (You can generate the data with some script you write). To profile your queries, learn how to read the results of the 2 following queries:
EXPLAIN extended...
SHOW STATUS
Then learn what to do to improve the figures (mysql settings, data structure, hardware, etc).
As a starter, I really advise these 2 great articles:
http://www.xaprb.com/blog/2006/10/12/how-to-profile-a-query-in-mysql/
http://ajohnstone.com/archives/mysql-php-performance-optimization-tips/
If you want to build for the academic beauty of the normalization: just follow the books and the general recommandations. :)
Out of the two options - like everyone before me said - it has to be option 1.
What you should really be concerned about are the bottlenecks in the application itself. Are users going to refer to the historical data often, or does it only show up for a few select users? If the answer is that everyone gets to see historical data of the ranks, then option 1 is good enough. If you are not going to refer to the historical ranks that often, then you could create a third "archive" table, and before updating the ranks, you can copy the rows of the original rank table to the archive table. This ensures that the number of rows stays minimal on the main table that is being called.
Remember, if you're updating the rows, and there's 10s of thousands, it might be more fruitful to get the results in your code (PHP/Python/etc), truncate the table and insert the results in rather than updating it row by row, as that would be a potential bottleneck.
You may want to look up sharding as well (horizontal partitioning) - http://en.wikipedia.org/wiki/Shard_%28database_architecture%29
And never forget to index well.
Hope that helped.
You stated the rank is only linked to the image, in which case all you need is table 1 and keep updating the rank in real time. Table 2 just stores unnecessary data. The disadvantage of this approach is that user cant change his vote.
You said the second table is for analysing /statistics, so it actually isn't something that needs to be stored in db. My suggestion is to get rid of the second table and use a logging facility to record rank changes.
Your second design is very dangerous in case you have 1 million votes for a photo. Can PHP handle that?
With the first design you can do all math on the database level which will be returning you a small result set.
Here is the scenario. 2 web servers in two separate locations having two mysql databases with identical tables. The data within the tables is also expected to be identical in real time.
Here is the problem. if a user in either location simultaneously enters a new record into identical tables, as illustrated in the two first tables below, where the third record in each table has been entered simultaneously by the different people. The data in the tables is no longer identical. Which is the best way to maintain that the data remains identical in real time as illustrated in the third table below regardless of where the updates take place? That way in the illustrations below instead of ending up with 3 rows in each table, the new records are replicated bi-directionally and they are inserted in both tables to create 2 identical tables again with 4 columns this time?
Server A in Location A
==============
Table Names
| ID| NAME |
|-----------|
| 1 | Tom |
| 2 | Scott |
|-----------|
| 3 | John |
|-----------|
Server B in Location B
==============
Table Names
| ID| NAME |
|-----------|
| 1 | Tom |
| 2 | Scott |
|-----------|
| 3 | Peter |
|-----------|
Expected Scenario
===========
Table Names
| ID| NAME |
|-----------|
| 1 | Tom |
| 2 | Scott |
| 3 | Peter |
| 4 | John |
|-----------|
There isn't much performance to be gained from replicating your database on two masters. However, there is a nifty bit of failover if you code your application correct.
Master-Master setup is essentially the same as the Slave-Master setup but has both Slaves started and an important change to your config files on each box.
Master MySQL 1:
auto_increment_increment = 2
auto_increment_offset = 1
Master MySQL 2:
auto_increment_increment = 2
auto_increment_offset = 2
These two parameters ensure that when two servers are fighting over a primary key for some reason, they do not duplicate and kill the replication. Instead of incrementing by 1, any auto-increment field will by default increment by 2. On one box it will start offset from 1 and run the sequence 1 3 5 7 9 11 13 etc. On the second box it will start offset at 2 and run along 2 4 6 8 10 12 etc. From current testing, the auto-increment appears to take the next free number, not one that has left before.
E.g. If server 1 inserts the first 3 records (1 3 and 5) when Server 2 inserts the 4th, it will be given the key of 6 (not 2, which is left unused).
Once you've set that up, start both of them up as Slaves.
Then to check both are working ok, connect to both machines and perform the command SHOW SLAVE STATUS and you should note that both Slave_IO_Running and Slave_SQL_Running should both say “YES” on each box.
Then, of course, create a few records in a table and ensure one box is only inserting odd numbered primary keys and the other is only incrementing even numbered ones.
Then do all the tests to ensure that you can perform all the standard applications on each box with it replicating to the other.
It's relatively simple once it's going.
But as has been mentioned, MySQL does discourage it and advise that you ensure you are mindful of this functionality when writing your application code.
Edit: I suppose it's theoretically possible to add more masters if you ensure that the offsets are correct and so on. You might more realistically though, add some additional slaves.
MySQL does not support synchronous replication, however, even if it did, you would probably not want to use it (can't take the performance hit of waiting for the other server to sync on every transaction commit).
You will have to consider more appropriate architectural solutions to it - there are third party products which will do a merge and resolve conflicts in a predetermined way - this is the only way really.
Expecting your architecture to function in this way is naive - there is no "easy fix" for any database, not just MySQL.
Is it important that the UIDs are the same? Or would you entertain the thought of having a table or column mapping the remote UID to the local UID and writing custom synchronisation code for objects you wish to replicate across that does any necessary mapping of UIDs for foreign key columns, etc?
The only way to ensure your tables are synchronized is to setup a 2-ways replication between databases.
But, MySQL only permits one-way replication, so you can't simply resolve your problem in this configuration.
To be clear, you can "setup" a 2-ways replication but MySQL AB discourages this.