Join on Apache Ignite REST API - rest

Does anyone know if you can do a join using the REST API for Apache Ignite? I have two objects, account and customer loaded to the Apache Ignite Server. Both objects are loaded with data and stored in the cache as account object cache and customer object cache. I am able to query both objects separately using the REST API, i.e.
http://localhost:8080/ignite?cmd=qryfldexe&pageSize=1000&cacheName=CustomerCache&qry=select+id+from+customer
http://localhost:8080/ignite?cmd=qryfldexe&pageSize=1000&cacheName=AccountCache&qry=select+id+from+account
However, I would like to execute a join on the account and customer cache. Is this supported and if so, does anyone have any examples? I can't find any documentation on this.

You need to specify one cache in the cacheName, and reference the second table in the JOIN via its schema name (by default it's the cache name). That's not unique to the REST API, Java API works in the same way.
The query should be something like
SELECT *
FROM Customer AS c
JOIN AccountCache.Account AS a
WHERE c.id = a.customerId

Try to use qryfldexe:
For example, I create next caches:
http://apache-ignite-users.70518.x6.nabble.com/file/t1704/ss1.java
It creates two caches with the same structure.
Now I am going to execute next command:
SELECT * FROM "mycache1".Value V1 join "mycache2".Value V2 on V1.key=V2.key
Let's use next converter to get URI string:
https://meyerweb.com/eric/tools/dencoder/
Our command will be next:
SELECT%20*%20FROM%20%22mycache1%22.Value%20V1%20join%20%22mycache2%22.Value%20V2%20on%20V1.key%3DV2.key
Run next in brouser:
http://127.0.0.1:8080/ignite?cmd=qryfldexe&pageSize=10&cacheName=mycache1&qry=SELECT%20*%20FROM%20%22mycache1%22.Value%20V1%20join%20%22mycache2%22.Value%20V2%20on%20V1.key%3DV2.key
Output:
{"successStatus":0,"error":null,"response":{"items":[[0,"Value 0",0,"Value
0"],[1,"Value 1",1,"Value 1"],[2,"Value 2",2,"Value 2"],[3,"Value
3",3,"Value 3"],[4,"Value 4",4,"Value 4"],[5,"Value 5",5,"Value
5"],[6,"Value 6",6,"Value 6"],[7,"Value 7",7,"Value 7"],[8,"Value
8",8,"Value 8"],[9,"Value 9",9,"Value
9"]],"last":false,"queryId":10,"fieldsMetadata":[{"schemaName":"mycache1","typeName":"VALUE","fieldName":"KEY","fieldTypeName":"java.lang.Integer"},{"schemaName":"mycache1","typeName":"VALUE","fieldName":"VALUE","fieldTypeName":"java.lang.String"},{"schemaName":"mycache2","typeName":"VALUE","fieldName":"KEY","fieldTypeName":"java.lang.Integer"},{"schemaName":"mycache2","typeName":"VALUE","fieldName":"VALUE","fieldTypeName":"java.lang.String"}]},"sessionToken":null}

Related

Spring JPA repository and QueryDsl how to force left join

Let's say I have two entities User and Task, each user can have one task.
The issue that I'm facing is if I have one record in the user table whose email starts with a and there are no records at all in the task table.
This snippet below will return no records although I would expect users that have mail starting with a.
UserRepository in example extends QuerydslPredicateExecutor.
userRepository.findAll(
QUser.user.email.startsWith("a")
.or(QUser.user.task.text.contains("something"))
)
If I check logs, Hibernate is creating cross join with user.task_id=task.id as a part of where clauses. This type of join automatically discards users whose mails are starting with a if they don't have a task assigned.
Is there a way to force usage of left join instead of a cross join in findAll method of the repository?
I know I can do it by using JPAQuery but then I would have to reimplement paging functionality...
JPAQuery query = new JPAQuery(entityManager);
query
.from(QUser.user)
.leftJoin(QTask.task)
// ...
I am not sure if we can do that since the findAll implementation is generated for us. However we can pass a predicate in the findAll method which will help deal with issue you are encountering.
You can try to do something like this:
QUser qUser = QUser.user;
QTask qTask = QTask.task;
JPQL<UserEntity> userJpqlQuery = JPAExpressions.selectFrom(qUser)
.leftjoin(qUser.task, qTask)
.where(qUser.email...., qTask.text...);
userRepository.findAll(qUser.in(userJpqlQuery));
In the code above I have used Querydsl, which is an alternative to CriteriaBuilder and is type safe. Then I have created a subquery to make the selection I want and return the all users matching the subquery.
In the end , hibernate should generate something like this:
select * from User qUser0 where qUser0.id.in(
select qUser1.id from User qUser1
left join Task qTask0 on
qUser1.taskId = qTask0.id
where ...
);

Is it possible to run a SQL query with EntityFramework that joins three tables between two databases?

So I've got a SQL query that is called from an API that I'm trying to write an integration test for. I have the method that prepares the data totally working, but I realized that I don't know how to actually execute the query to check that data (and run the test). Here is what the query looks like (slightly redacted to protect confidental data):
SELECT HeaderQuery.[headerid],
kaq.[applicationname],
HeaderQuery.[usersession],
HeaderQuery.[username],
HeaderQuery.[referringurl],
HeaderQuery.[route],
HeaderQuery.[method],
HeaderQuery.[logdate],
HeaderQuery.[logtype],
HeaderQuery.[statuscode],
HeaderQuery.[statusdescription],
DetailQuery.[detailid],
DetailQuery.[name],
DetailQuery.[value]
FROM [DATABASE1].[dbo].[apilogheader] HeaderQuery
LEFT JOIN [DATABASE1].[dbo].[apilogdetails] DetailQuery
ON HeaderQuery.[headerid] = DetailQuery.[headerid]
INNER JOIN [DATABASE2].[dbo].[apps] kaq
ON HeaderQuery.[applicationid] = kaq.[applicationid]
WHERE HeaderQuery.[applicationid] = #applicationid1
AND HeaderQuery.[logdate] >= #logdate2
AND HeaderQuery.[logdate] <= #logdate3
For the sake of the test, and considering I already have the SQL script, I was hoping to be able to just execute that script above (providing the where clause programmatically) using context.Database.SqlQuery<string>(QUERY) but since I have two different contexts, I'm not sure how to do that.
The short answer is no, EF doesn’t support cross database queries. However there are a few things you can try.
You can use two different database contexts (one for each database).
Run your respective queries and then merge / massage the data after
the query returns.
Create a database view and query the view through EF.
Using a SYNONYM
https://rachel53461.wordpress.com/2011/05/22/tricking-ef-to-span-multiple-databases/
If the databases are on the same server, you can try using a
DbCommandInterceptor
I’ve had this requirement before and personally like the view option.

Spring CRUD repository: is there findOneByMaxXYZColumn()?

My requirement:
fetch ONE object (e.g RetainInfo ) from table RETAIN_INFO if VERSION column has max value
Does CRUD repository support for an interface method like
findOneByMaxRetVersionAndCountry("DEFAULT")
Equivalent db2 sql:
select RET_ID, max(ri.RET_VERSION) from RETAIN_INFO ri where ri. COUNTRY='DEFAULT' group by RET_ID fetch first 1 rows only;
This query selects an ID, but I would actually want the RetainInfo object corresponding the SINGLE row returned by the query.
I prefer to get that without using custom query, i.e using findBy or some other method/interface supported by Spring CRUD.
You could use limiting in combination with sorting (spring data reference:limit query results). Declare a method similar to the following in your CrudRepository interface :
RetainInfo findTopByCountryOrderByRetVersionDesc(String country);
You can also use findFirst to get the first result. Before getting the result, make sure to use Orderby and then the ascending(Asc) or descending(Desc). As an example if you want to order by version and retrieve based on productName
RetainInfo findFirstByProductNameOrderByVersionDesc(String productName);
Spring Data doesn't provide an expression to select a max value. All supported query parts could be found in the Spring 1.2.0.RELEASE docs: Appendix A. Namespace reference or line 182 of org.springframework.data.repository.query.parser.Part.
Also feel free to create a feature request at Spring's Jira page.

Tableau Extract API with multiple tables in a database

I am currently experimenting with Tableau Extract API to generate some TDE from the tables I have in a PostgreSQL database. I was able to write a code to generate the TDE from single table, but I would like to do this for multiple joined tables. To be more specific, if I have two tables that are inner joined by some field, how would I generate the TDE for this?
I can see that if I am working with small number of tables, I could use a SQL query with JOIN clauses to create a one gigantic table, and generate the TDE from that table.
>> SELECT * FROM table_1 INNER JOIN table_2
INTO new_table_1
ON table_1.id_1 = table_2.id_2;
>> SELECT * FROM new_table_1 INNER JOIN TABLE_3
INTO new_table_2
ON new_table_1.id_1 = table_3.id_3
and then generate the TDE from new_table_2.
However, I have some tables that have over 40 different fields, so this could get messy.
Is this even a possibility with current version of the API?
You can read from as many tables or other sources as you want. Or use complex query with lots of joins, or create a view and read from that. Usually, creating a view is helpful when you have a complex query joining many tables.
The data extract API is totally agnostic about how or where you get the data to feed it -- the whole point is to allow you to grab data from unusual sources that don't have pre-built drivers for Tableau.
Since Tableau has a Postgres driver and can read from it directly, you don't need to write a program with the data extract API at all. You can define your extract with Tableau Desktop. If you need to schedule automated refreshes of the extract, you can use Tableau Server or its tabcmd command.
Many thanks for your replies. I am aware that I could use Tableau Desktop to define my extract. In fact, I have done this many times before. I am just trying to create the extracts using the API, because I need to create some calculated fields, which is near impossible to create using the Tableau Desktop.
At this point, I am hesitant to use JOINs in the SQL query because the resulting table would look too complicated to comprehend (some of these tables also have same field names).
When you say that I could read from multiple tables or sources, does that mean with the Tableau Extract API? At this point, I cannot find anywhere in this API that accommodates multiple sources. For example, I know that when I use multiple tables in the Tableau Desktop, there are icons on the left hand side that tells me that the extract is composed of multiple tables. This just doesn't seem to be happening with the API, which leaves me stranded. Anyways, thank you again for your replies.
Going back to the topic, this is something that I tried few days ago on my python code
try:
tdefile= tde.Extract("extract.tde")
except:
os.remove("extract.tde")
tdefile = tde.Extract("extract.tde")
tableDef = tde.TableDefinition()
# Read each column in table and set the column data types using tableDef.addColumn
# Some code goes here...
for eachTable in tableNames:
tableAdd = tdeFile.addTable(eachTable, tableDef)
# Use SQL query to retrieve bunch_of_rows from eachTable
for some_row in bunch_of_rows:
# Read each row in table, and set the values in each column position of each row
# Some code goes here...
tableAdd.insert(some_row)
some_row.close()
tdefile.close()
When I execute this code, I get the error that eachTable has to be called "Extract".
Of course, this code has its flaws, as there is no where in this code that tells how each table are being joined.
So I am little thrown off here, because it doesn't seem like I can use multiple tables unless I use JOINs to generate one table that contains everything.

How to build a select using Zend with a DISTINCT specific column?

I'm using Zend Framework for my website and I'd like to retrieve some data from my PostgreSQL database.
I have a request like :
SELECT DISTINCT ON(e.id) e.*, f.*, g.* FROM e, f, g
WHERE e.id = f.id_e AND f.id = g.id_f
This request works well but I don't know how to convert the DISTINCT ON(e.id) with Zend.
It seems that I can get DISTINCT rows but no distinct columns.
$select->distinct()->from("e")->join("f", "e.id = f.id_e")
->join("g", "f.id = g.id_f");
Any idea on how to make a select with distinct column ?
Thanks for help
You probably can't do this with Zend Framework since distinct on is not part of the SQL standard (end of page in Postgres documentation). Although Postgres supports it, I would assume its not part of Zend Framework because you could in theory configure another database connection which does not offer support.
If you know in advance that you're developing for a specific database (Postgres in this case), you could use manually written statements instead. You'll gain more flexibility within the queries and better performance at the cost of no longer being able to switch databases.
You would then instantiate a Zend_Db_Apdapter for Postgres. There a various methods available to get results for SQL queries which are described in the frameworks documentation starting at section Reading Query Results. If you choose to go this route I'd recommend to create an own subclass of the Zend_Db_Adapter_Pgsql class. This is to be able to convert data types and throw exceptions in case of errors instead of returning ambiguous null values and hiding error causes.