You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@ignite.apache.org by Shai N <sn...@jobcase.com> on 2018/05/18 17:38:36 UTC
QueryEntity not reflecting ALTER TABLE changes made via DDL
I'm trying to detect in Java which fields in a cache are exposed to SQL by
looking at
"((QueryEntity)ignite.cache("cacheName").getConfiguration(CacheConfiguration.class).getQueryEntities().iterator().next()).getFields()".
This accurately reflects a table's initial schema. However, changes made
with "ALTER TABLE ADD COLUMN" and "DROP COLUMN" are not reflected in the
QueryEntity. For example, if I create then alter a table via a SQL client
with:
---
CREATE TABLE test1 (
"key" VARCHAR(30) NOT NULL,
"id" BIGINT(20) NOT NULL,
"testColumn1" VARCHAR(24) NOT NULL,
PRIMARY KEY("key"))
WITH "affinitykey=key, backups=1, write_synchronization_mode=FULL_SYNC,
atomicity=transactional, cache_name=test1, value_type=test1,
key_type=arrivals_key";
ALTER TABLE test1 ADD COLUMN "testColumn2" VARCHAR(24) NOT NULL;
---
... then try to print out the result of
"((QueryEntity)ignite.cache("test1").getConfiguration(CacheConfiguration.class).getQueryEntities().iterator().next()).getFields()"
I get "{key=java.lang.String, id=java.lang.Long,
testColumn1=java.lang.String}" which doesn't include the new column.
Dropping columns is likewise not reflected.
When I do try to query the cache via SQL, the added and dropped columns are
reflected in the result, so it's clear the change didn't fail to go through
entirely.
Is there something I'm missing? Is it possible to alter a table and then
detect the column changes in Java without entirely recreating the cache?
--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/
Re: QueryEntity not reflecting ALTER TABLE changes made via DDL
Posted by Ilya Kasnacheev <il...@gmail.com>.
Hello!
Unfortunately there's no support in Native SQL API for table introspection.
You could use thin client since it's fairly lighweight.
Regards,
--
Ilya Kasnacheev
2018-05-24 23:50 GMT+03:00 Nick Pordash <ni...@gmail.com>:
> Hi Illya,
>
> Is this possible to get without having to resort to JDBC? For example, the
> Ignite SQL Java Apis?
>
> -Nick
>
>
> On Mon, May 21, 2018, 5:56 AM Ilya Kasnacheev <il...@gmail.com>
> wrote:
>
>> Hello!
>>
>> Cache Configuration would only request its configuration at the time when
>> it was started. It will not be modified when DDL is issued.
>>
>> Yes, you can detect column changes, e.g. by using JDBC Metadata:
>> https://docs.oracle.com/javase/7/docs/api/java/sql/DatabaseMetaData.html
>>
>> Regards,
>>
>> --
>> Ilya Kasnacheev
>>
>> 2018-05-18 20:38 GMT+03:00 Shai N <sn...@jobcase.com>:
>>
>>> I'm trying to detect in Java which fields in a cache are exposed to SQL
>>> by
>>> looking at
>>> "((QueryEntity)ignite.cache("cacheName").getConfiguration(
>>> CacheConfiguration.class).getQueryEntities().iterator().
>>> next()).getFields()".
>>> This accurately reflects a table's initial schema. However, changes made
>>> with "ALTER TABLE ADD COLUMN" and "DROP COLUMN" are not reflected in the
>>> QueryEntity. For example, if I create then alter a table via a SQL client
>>> with:
>>> ---
>>> CREATE TABLE test1 (
>>> "key" VARCHAR(30) NOT NULL,
>>> "id" BIGINT(20) NOT NULL,
>>> "testColumn1" VARCHAR(24) NOT NULL,
>>> PRIMARY KEY("key"))
>>> WITH "affinitykey=key, backups=1, write_synchronization_mode=FULL_SYNC,
>>> atomicity=transactional, cache_name=test1, value_type=test1,
>>> key_type=arrivals_key";
>>>
>>> ALTER TABLE test1 ADD COLUMN "testColumn2" VARCHAR(24) NOT NULL;
>>> ---
>>> ... then try to print out the result of
>>> "((QueryEntity)ignite.cache("test1").getConfiguration(
>>> CacheConfiguration.class).getQueryEntities().iterator().
>>> next()).getFields()"
>>> I get "{key=java.lang.String, id=java.lang.Long,
>>> testColumn1=java.lang.String}" which doesn't include the new column.
>>> Dropping columns is likewise not reflected.
>>>
>>> When I do try to query the cache via SQL, the added and dropped columns
>>> are
>>> reflected in the result, so it's clear the change didn't fail to go
>>> through
>>> entirely.
>>>
>>> Is there something I'm missing? Is it possible to alter a table and then
>>> detect the column changes in Java without entirely recreating the cache?
>>>
>>>
>>>
>>> --
>>> Sent from: http://apache-ignite-users.70518.x6.nabble.com/
>>>
>>
>>
Re: QueryEntity not reflecting ALTER TABLE changes made via DDL
Posted by Nick Pordash <ni...@gmail.com>.
Hi Illya,
Is this possible to get without having to resort to JDBC? For example, the
Ignite SQL Java Apis?
-Nick
On Mon, May 21, 2018, 5:56 AM Ilya Kasnacheev <il...@gmail.com>
wrote:
> Hello!
>
> Cache Configuration would only request its configuration at the time when
> it was started. It will not be modified when DDL is issued.
>
> Yes, you can detect column changes, e.g. by using JDBC Metadata:
> https://docs.oracle.com/javase/7/docs/api/java/sql/DatabaseMetaData.html
>
> Regards,
>
> --
> Ilya Kasnacheev
>
> 2018-05-18 20:38 GMT+03:00 Shai N <sn...@jobcase.com>:
>
>> I'm trying to detect in Java which fields in a cache are exposed to SQL by
>> looking at
>>
>> "((QueryEntity)ignite.cache("cacheName").getConfiguration(CacheConfiguration.class).getQueryEntities().iterator().next()).getFields()".
>> This accurately reflects a table's initial schema. However, changes made
>> with "ALTER TABLE ADD COLUMN" and "DROP COLUMN" are not reflected in the
>> QueryEntity. For example, if I create then alter a table via a SQL client
>> with:
>> ---
>> CREATE TABLE test1 (
>> "key" VARCHAR(30) NOT NULL,
>> "id" BIGINT(20) NOT NULL,
>> "testColumn1" VARCHAR(24) NOT NULL,
>> PRIMARY KEY("key"))
>> WITH "affinitykey=key, backups=1, write_synchronization_mode=FULL_SYNC,
>> atomicity=transactional, cache_name=test1, value_type=test1,
>> key_type=arrivals_key";
>>
>> ALTER TABLE test1 ADD COLUMN "testColumn2" VARCHAR(24) NOT NULL;
>> ---
>> ... then try to print out the result of
>>
>> "((QueryEntity)ignite.cache("test1").getConfiguration(CacheConfiguration.class).getQueryEntities().iterator().next()).getFields()"
>> I get "{key=java.lang.String, id=java.lang.Long,
>> testColumn1=java.lang.String}" which doesn't include the new column.
>> Dropping columns is likewise not reflected.
>>
>> When I do try to query the cache via SQL, the added and dropped columns
>> are
>> reflected in the result, so it's clear the change didn't fail to go
>> through
>> entirely.
>>
>> Is there something I'm missing? Is it possible to alter a table and then
>> detect the column changes in Java without entirely recreating the cache?
>>
>>
>>
>> --
>> Sent from: http://apache-ignite-users.70518.x6.nabble.com/
>>
>
>
Re: QueryEntity not reflecting ALTER TABLE changes made via DDL
Posted by Ilya Kasnacheev <il...@gmail.com>.
Hello!
Cache Configuration would only request its configuration at the time when
it was started. It will not be modified when DDL is issued.
Yes, you can detect column changes, e.g. by using JDBC Metadata:
https://docs.oracle.com/javase/7/docs/api/java/sql/DatabaseMetaData.html
Regards,
--
Ilya Kasnacheev
2018-05-18 20:38 GMT+03:00 Shai N <sn...@jobcase.com>:
> I'm trying to detect in Java which fields in a cache are exposed to SQL by
> looking at
> "((QueryEntity)ignite.cache("cacheName").getConfiguration(
> CacheConfiguration.class).getQueryEntities().iterator().
> next()).getFields()".
> This accurately reflects a table's initial schema. However, changes made
> with "ALTER TABLE ADD COLUMN" and "DROP COLUMN" are not reflected in the
> QueryEntity. For example, if I create then alter a table via a SQL client
> with:
> ---
> CREATE TABLE test1 (
> "key" VARCHAR(30) NOT NULL,
> "id" BIGINT(20) NOT NULL,
> "testColumn1" VARCHAR(24) NOT NULL,
> PRIMARY KEY("key"))
> WITH "affinitykey=key, backups=1, write_synchronization_mode=FULL_SYNC,
> atomicity=transactional, cache_name=test1, value_type=test1,
> key_type=arrivals_key";
>
> ALTER TABLE test1 ADD COLUMN "testColumn2" VARCHAR(24) NOT NULL;
> ---
> ... then try to print out the result of
> "((QueryEntity)ignite.cache("test1").getConfiguration(
> CacheConfiguration.class).getQueryEntities().iterator().
> next()).getFields()"
> I get "{key=java.lang.String, id=java.lang.Long,
> testColumn1=java.lang.String}" which doesn't include the new column.
> Dropping columns is likewise not reflected.
>
> When I do try to query the cache via SQL, the added and dropped columns are
> reflected in the result, so it's clear the change didn't fail to go through
> entirely.
>
> Is there something I'm missing? Is it possible to alter a table and then
> detect the column changes in Java without entirely recreating the cache?
>
>
>
> --
> Sent from: http://apache-ignite-users.70518.x6.nabble.com/
>