I have a few tables with about 17M rows that all have a date column I would like to be able to utilize frequently for searches. I am considering either just throwing an index on the column and see how things go or sorting the items by date as a one time operation and then inserting everything into a new table so that the primary key ascends as the date ascends.
Since these are both pretty time consuming I thought it might be worth it to ask here first for input.
The end goal is for me to load sql queries into pandas for some analysis if that is relevant here.
The index on a date column makes sense when you are going to search the table for a given date(s), e.g.:
select * from test
where the_date = '2016-01-01';
-- or
select * from test
where the_date between '2016-01-01' and '2016-01-31';
-- etc
In these queries there is no matter whether the sort order of primary key and the date column are the same or not. Hence rewriting the data to the new table will be useless. Just create an index.
However, if you are going to use the index only in ORDER BY:
select * from test
order by the_date;
then a primary key integer index may be significantly (2-4 times) faster then an index on a date column.
Postgres supports to some extend clustered indexes, which is what you suggest by removing and reinserting the data.
In fact, removing and reinserting the data in the order you want will not change the time the query takes. Postgres does not know the order of the data.
If you know that the table's data does not change. Then cluster the data based on the index you create.
This operation reorders the table based on the order in the index. It is very effective until you update the table. The syntax is:
CLUSTER tableName USING IndexName;
See the manual for details.
I also recommend you use
explain <query>;
to compare two queries, before and after an index. Or before and after clustering.
Related
I've recently started to play around with Cassandra. My understanding is that in a Cassandra table you define 2 keys, which can be either single column or composites:
The Partitioning Key: determines how to distribute data across nodes
The Clustering Key: determines in which order the records of a same partitioning key (i.e. within a same node) are written. This is also the order in which the records will be read.
Data from a table will always be sorted in the same order, which is the order of the clustering key column(s). So a table must be designed for a specific query.
But what if I need to perform 2 different queries on the data from a table. What is the best way to solve this when using Cassandra ?
Example Scenario
Let's say I have a simple table containing posts that users have written :
CREATE TABLE posts (
username varchar,
creation timestamp,
content varchar,
PRIMARY KEY ((username), creation)
);
This table was "designed" to perform the following query, which works very well for me:
SELECT * FROM posts WHERE username='luke' [ORDER BY creation DESC];
Queries
But what if I need to get all posts regardless of the username, in order of time:
Query (1): SELECT * FROM posts ORDER BY creation;
Or get the posts in alphabetical order of the content:
Query (2): SELECT * FROM posts WHERE username='luke' ORDER BY content;
I know that it's not possible given the table I created, but what are the alternatives and best practices to solve this ?
Solution Ideas
Here are a few ideas spawned from my imagination (just to show that at least I tried):
Querying with the IN clause to select posts from many users. This could help in Query (1). When using the IN clause, you can fetch globally sorted results if you disable paging. But using the IN clause quickly leads to bad performance when the number of usernames grows.
Maintaining full copies of the table for each query, each copy using its own PRIMARY KEY adapted to the query it is trying to serve.
Having a main table with a UUID as partitioning key. Then creating smaller copies of the table for each query, which only contain the (key) columns useful for their own sort order, and the UUID for each row of the main table. The smaller tables would serve only as "sorting indexes" to query a list of UUID as result, which can then be fetched using the main table.
I'm new to NoSQL, I would just want to know what is the correct/durable/efficient way of doing this.
The SELECT * FROM posts ORDER BY creation; will results in a full cluster scan because you do not provide any partition key. And the ORDER BY clause in this query won't work anyway.
Your requirement I need to get all posts regardless of the username, in order of time is very hard to achieve in a distributed system, it supposes to:
fetch all user posts and move them to a single node (coordinator)
order them by date
take top N latest posts
Point 1. require a full table scan. Indeed as long as you don't fetch all records, the ordering can not be achieve. Unless you use Cassandra clustering column to order at insertion time. But in this case, it means that all posts are being stored in the same partition and this partition will grow forever ...
Query SELECT * FROM posts WHERE username='luke' ORDER BY content; is possible using a denormalized table or with the new materialized view feature (http://www.doanduyhai.com/blog/?p=1930)
Question 1:
Depending on your use case I bet you could model this with time buckets, depending on the range of times you're interested in.
You can do this by making the primary key a year,year-month, or year-month-day depending on your use case (or finer time intervals)
The basic idea is that you bucket changes for what suites your use case. For example:
If you often need to search these posts over months in the past, then you may want to use the year as the PK.
If you usually need to search the posts over several days in the past, then you may want to use a year-month as the PK.
If you usually need to search the post for yesterday or a couple of days, then you may want to use a year-month-day as your PK.
I'll give a fleshed out example with yyyy-mm-dd as the PK:
The table will now be:
CREATE TABLE posts_by_creation (
creation_year int,
creation_month int,
creation_day int,
creation timeuuid,
username text, -- using text instead of varchar, they're essentially the same
content text,
PRIMARY KEY ((creation_year,creation_month,creation_day), creation)
)
I changed creation to be a timeuuid to guarantee a unique row for each post creation event. If we used just a timestamp you could theoretically overwrite an existing post creation record in here.
Now we can then insert the Partition Key (PK): creation_year, creation_month, creation_day based on the current creation time:
INSERT INTO posts_by_creation (creation_year, creation_month, creation_day, creation, username, content) VALUES (2016, 4, 2, now() , 'fromanator', 'content update1';
INSERT INTO posts_by_creation (creation_year, creation_month, creation_day, creation, username, content) VALUES (2016, 4, 2, now() , 'fromanator', 'content update2';
now() is a CQL function to generate a timeUUID, you would probably want to generate this in the application instead, and parse out the yyyy-mm-dd for the PK and then insert the timeUUID in the clustered column.
For a usage case using this table, let's say you wanted to see all of the changes today, your CQL would look like:
SELECT * FROM posts_by_creation WHERE creation_year = 2016 AND creation_month = 4 AND creation_day = 2;
Or if you wanted to find all of the changes today after 5pm central:
SELECT * FROM posts_by_creation WHERE creation_year = 2016 AND creation_month = 4 AND creation_day = 2 AND creation >= minTimeuuid('2016-04-02 5:00-0600') ;
minTimeuuid() is another cql function, it will create the smallest possible timeUUID for the given time, this will guarantee that you get all of the changes from that time.
Depending on the time spans you may need to query a few different partition keys, but it shouldn't be that hard to implement. Also you would want to change your creation column to a timeuuid for your other table.
Question 2:
You'll have to create another table or use materialized views to support this new query pattern, just like you thought.
Lastly if your not on Cassandra 3.x+ or don't want to use materialized views you can use Atomic batches to ensure data consistency across your several de-normalized tables (that's what it was designed for). So in your case it would be a BATCH statement with 3 inserts of the same data to 3 different tables that support your query patterns.
The solution is to create another tables to support your queries.
For SELECT * FROM posts ORDER BY creation;, you may need some special column for grouping it, maybe by month and year, e.g. PRIMARY KEY((year, month), timestamp) this way the cassandra will have a better performance on read because it doesn't need to scan the whole cluster to get all data, it will also save the data transfer between nodes too.
Same as SELECT * FROM posts WHERE username='luke' ORDER BY content;, you must create another table for this query too. All column may be same as your first table but with the different Primary Key, because you cannot order by the column that is not the clustering column.
Consider this Oracle docs about indexes, this about speed of insert and this question on StackOverflow lead me to conclusion that:
Indexes helps us locate information faster
Primary and Unique Keys are indexed automatically
Inserting with indexes can cause worse performance
However every time indexes are discussed there are only SELECT operations shown as examples.
My question is: are indexes used in INSERT and UPDATE operations? When and how?
My suggestions are:
UPDATE can use index in WHERE clause (if the column in the clause has index)
INSERT can use index when uses SELECT (but in this case, index is from another table)
or probably when checking integrity constraints
but I don't have such deep knowledge of using indexes.
For UPDATE statements, index can be used by the optimiser if it deems the index can speed it up. The index would be used to locate the rows to be updated. The index is also a table in a manner of speaking, so if the indexed column is getting updated, it obviously needs to UPDATE the index as well. On the other hand if you're running an update without a WHERE clause the optimiser may choose not to use an index as it has to access the whole table, a full table scan may be more efficient (but may still have to update the index). The optimiser makes those decisions at runtime based on several parameters such as if there are valid stats against the tables and indexes in question, how much data is affected, what type of hardware, etc.
For INSERT statements though the INSERT itself does not need the index, the index will also need to be 'inserted into', so will need to be accessed by oracle. Another case where INSERT can cause the index to be used is an INSERT like this:
INSERT INTO mytable (mycolmn)
SELECT mycolumn + 10 FROM mytable;
Insert statement has no direct benefit for index. But more index on a table cause slower insert operation. Think about a table that has no index on it and if you want to add a row on it, it will find table block that has enough free space and store that row. But if that table has indexes on it database must make sure that these new rows also found via indexes, So to add new rows on a table that has indexes, also need to entry in indexes too. That multiplies the insert operation. So more index you have, more time you need to insert new rows.
For update it depends on whether you update indexed column or not. If you are not updating indexed column then performance should not be affected. Index can also speed up a update statements if the where conditions can make use of indexes.
I have a large table of 7 column in Oracle 11G. Total size of the table is more than 3GB and total row in this table is 1876823. Query we are using
select doc_mstr_id from index_mstr where page_con1 like('%sachin%') it is taking almost a minute. please help me to optimize the query as well as proper indexing for this table. Please let me know if partitioned is required for this table.
Below are the column description
INDEX_MSTR_ID NUMBER
DOC_MSTR_ID NUMBER
PAGE_NO NUMBER
PAGE_PART NUMBER
PAGE_CON1 VARCHAR2(4000)
FILE_MODIFIED_DATE DATE
CREATED_DATE DATE
This query is always going to result in a full table scan. Your only filter cannot use a B-TREE index, due to the leading wildcard:
where page_con1 like('%sachin%')
If you want to do lots of queries of this nature you need to build a Text index on that column. From its datatype page_con1 appears to hold text fragments rather than full documents so you should use a CTXCAT index. This type of index has the advantage of being transactional, rather than requiring background maintenance. Find out more.
Your query would then look like this:
select doc_mstr_id from index_mstr
WHERE CATSEARCH(page_con1, 'sachin') > 0;
In our one of production database, we have 4 column table and there are no PK,UK constraints on it. only one notnull constraint on one column. The inserts are slow on this table and when I checked the indexes , there is one index which is built on all columns.
It is a normal table and not IOT. I really don't see a need of all column index, but wondering why the developers has created it?
Appreciate your thoughts?
It might be usefull, i.e. if you (mainly) query all columns oracle doesn't have to access the table at all, but can get all the data from the index. Though inserts take longer because a larger index has to be maintained by the dbms everytime.
One case where it could be useful is,
Say for example, you are trying to check the existence of records in this table and for that you have to have joins on all four columns. So in such a case if you have written a correlated query like below,
SELECT <something>
FROM table_1 t1
WHERE EXISTS
(SELECT 1 FROM table_t2 t2 where t1.c1=t2.c1 and t1.c2=t2.c2 and t1.c3=t2.c3 and t1.c4=t2.c4)
Apart from above case, it looks an error to me from developer's side.
Indexes are good to better query optimization but causes slow updates/inserts because the indexes needs to be updated at each modification.
If these tables first use is querying and inserts happens only in a specific periods like a batch at the beginning or the end of the day only, then you can remove the indexes before updating tables and then restore them.
In addition, all the queries all these tables need to be analysed to see which indexes are useful and which are not?
Anyway, You need to ask developers before removing these indexes.
I have a table that I expect to get 7 million records a month on a pretty wide table. A small portion of these records are expected to be flagged as "problem" records.
What is the best way to implement the table to locate these records in an efficient way?
I'm new to Oracle, but is a materialized view an valid option? Are there such things in Oracle such as indexed views or is this potentially really the same thing?
Most of the reporting is by month, so partitioning by month seems like an option, but a "problem" record may be lingering for several months theorectically. Otherwise, the reporting shuold be mostly for the current month. Would you expect that querying across all month partitions to locate any problem record would cause significant performance issues compared to usinga single table?
Your general thoughts of where to start would be appreciated. I realize I need to read up and I'll do that but I wanted to get the community thought first to make sure I read the right stuff.
One more thought: The primary key is a GUID varchar2(36). In order of magnitude, how much of a performance hit would you expect this to be relative to using a NUMBER data type PK? This worries me but it is out of my control.
It depends what you mean by "flagged", but it sounds to me like you would benefit from a simple index, function based index, or an indexed virtual column.
In all cases you should be careful to ensure that all the index columns are NULL for rows that do not need to be flagged. This way your index will contain only the rows that are flagged (Oracle does not - by default - index rows in B-Tree indexes where all index column values are NULL).
Your primary key being a VARCHAR2 GUID should make no difference, at least with regards to the specific flagging of rows in this question, indexes will point to rows via Oracle internal ROWIDs.
Indexes support partitioning, so if your data is already partitioned, your index could be set to match.
Simple column index method
If you can dictate how the flagging works, or the column already exists, then I would simply add an index to it like so:
CREATE INDEX my_table_problems_idx ON my_table (problem_flag)
/
Function-based index method
If the data model is fixed / there is no flag column, then you can create a function-based index assuming that you have all the information you need in the target table. For example:
CREATE INDEX my_table_problems_fnidx ON my_table (
CASE
WHEN amount > 100 THEN 'Y'
ELSE NULL
END
)
/
Now if you use the same logic in your SELECT statement, you should find that it uses the index to efficiently match rows.
SELECT *
FROM my_table
WHERE CASE
WHEN amount > 100 THEN 'Y'
ELSE NULL
END IS NOT NULL
/
This is a bit clunky though, and it requires you to use the same logic in queries as the index definition. Not great. You could use a view to mask this, but you're still duplicating logic in at least two places.
Indexed virtual column
In my opinion, this is the best way to do it if you are computing the value dynamically (available from 11g onwards):
ALTER TABLE my_table
ADD virtual_problem_flag VARCHAR2(1) AS (
CASE
WHEN amount > 100 THEN 'Y'
ELSE NULL
END
)
/
CREATE INDEX my_table_problems_idx ON my_table (virtual_problem_flag)
/
Now you can just query the virtual column as if it were a real column, i.e.
SELECT *
FROM my_table
WHERE virtual_problem_flag = 'Y'
/
This will use the index and puts the function-based logic into a single place.
Create a new table with just the pks of the problem rows.