You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@drill.apache.org by "Jaimes, Rafael - 0993 - MITLL" <Ra...@ll.mit.edu> on 2020/04/01 16:45:36 UTC
query Presto via Drill
As a proof of concept, I thought it might be interesting to see if Drill
could query Presto as if it were a RDBMS.
In theory, I thought it would work because it uses a JDBC driver. I plopped
it into 3rdparty and setup the config (see bottom) and it saves without
error. However it doesn't show up in 'show schemas;' All attempts at
querying it or showing tables fail.
Anyone else crazy enough to attempt this?
If you were wondering, I am interested in a connector that Presto has that
Drill doesn't, but Drill has better support for other things so I'd like to
keep it on top.
{
"type": "jdbc",
"driver": "io.prestosql.jdbc.PrestoDriver",
"url": "jdbc:presto://presto-url:9999/catalog/schema",
"username": "user",
"password": null,
"caseInsensitiveTableNames": false,
"sourceParameters": {},
"enabled": true
}
Re: query Presto via Drill
Posted by Charles Givre <cg...@gmail.com>.
Hey Rafael,
Regarding the earlier issue you were having with Drill and Presto, I opened a JIRA about this. My concern is that the JDBC plugin might have a bug that needs some attention. Here's the JIRA if you want to add anything:
https://issues.apache.org/jira/browse/DRILL-7698 <https://issues.apache.org/jira/browse/DRILL-7698>
-- C
> On Apr 1, 2020, at 1:29 PM, Rafael Jaimes III <ra...@gmail.com> wrote:
>
> Apache Accumulo. I think it might be possible to setup a Hive metastore and
> run it though that - but we don't have any Hive instances setup and I'm
> just trying things out with current infrastructure. Somebody else already
> setup Accumulo+Presto.
>
> I found a couple of inquiries regarding Accumulo plugin for Drill in the
> list archive, but there doesn't seem to be a suitable option yet. Working
> off HBase plugin could eventually be a route but I don't have the manpower
> or expertise at the moment to go down that road.
>
> Thanks,
> Rafael
>
> On Wed, Apr 1, 2020, 1:17 PM Charles Givre <cgivre@gmail.com <ma...@gmail.com>> wrote:
>
>> Hi Rafael,
>> What connector are you interested in specifically?
>> Thanks,
>> -- C
>>
>>
>>> On Apr 1, 2020, at 12:45 PM, Jaimes, Rafael - 0993 - MITLL <
>> Rafael.Jaimes@ll.mit.edu> wrote:
>>>
>>> As a proof of concept, I thought it might be interesting to see if Drill
>> could query Presto as if it were a RDBMS.
>>> In theory, I thought it would work because it uses a JDBC driver. I
>> plopped it into 3rdparty and setup the config (see bottom) and it saves
>> without error. However it doesn’t show up in ‘show schemas;’ All attempts
>> at querying it or showing tables fail.
>>> Anyone else crazy enough to attempt this?
>>> If you were wondering, I am interested in a connector that Presto has
>> that Drill doesn’t, but Drill has better support for other things so I’d
>> like to keep it on top.
>>>
>>> {
>>> "type": "jdbc",
>>> "driver": "io.prestosql.jdbc.PrestoDriver",
>>> "url": "jdbc:presto://presto-url:9999/catalog/schema
>> <presto://presto-url:9999/catalog/schema <presto://presto-url:9999/catalog/schema>>",
>>> "username": "user",
>>> "password": null,
>>> "caseInsensitiveTableNames": false,
>>> "sourceParameters": {},
>>> "enabled": true
>>> }
Re: query Presto via Drill
Posted by Rafael Jaimes III <ra...@gmail.com>.
Apache Accumulo. I think it might be possible to setup a Hive metastore and
run it though that - but we don't have any Hive instances setup and I'm
just trying things out with current infrastructure. Somebody else already
setup Accumulo+Presto.
I found a couple of inquiries regarding Accumulo plugin for Drill in the
list archive, but there doesn't seem to be a suitable option yet. Working
off HBase plugin could eventually be a route but I don't have the manpower
or expertise at the moment to go down that road.
Thanks,
Rafael
On Wed, Apr 1, 2020, 1:17 PM Charles Givre <cg...@gmail.com> wrote:
> Hi Rafael,
> What connector are you interested in specifically?
> Thanks,
> -- C
>
>
> > On Apr 1, 2020, at 12:45 PM, Jaimes, Rafael - 0993 - MITLL <
> Rafael.Jaimes@ll.mit.edu> wrote:
> >
> > As a proof of concept, I thought it might be interesting to see if Drill
> could query Presto as if it were a RDBMS.
> > In theory, I thought it would work because it uses a JDBC driver. I
> plopped it into 3rdparty and setup the config (see bottom) and it saves
> without error. However it doesn’t show up in ‘show schemas;’ All attempts
> at querying it or showing tables fail.
> > Anyone else crazy enough to attempt this?
> > If you were wondering, I am interested in a connector that Presto has
> that Drill doesn’t, but Drill has better support for other things so I’d
> like to keep it on top.
> >
> > {
> > "type": "jdbc",
> > "driver": "io.prestosql.jdbc.PrestoDriver",
> > "url": "jdbc:presto://presto-url:9999/catalog/schema
> <presto://presto-url:9999/catalog/schema>",
> > "username": "user",
> > "password": null,
> > "caseInsensitiveTableNames": false,
> > "sourceParameters": {},
> > "enabled": true
> > }
>
>
Re: query Presto via Drill
Posted by Charles Givre <cg...@gmail.com>.
Hi Rafael,
What connector are you interested in specifically?
Thanks,
-- C
> On Apr 1, 2020, at 12:45 PM, Jaimes, Rafael - 0993 - MITLL <Ra...@ll.mit.edu> wrote:
>
> As a proof of concept, I thought it might be interesting to see if Drill could query Presto as if it were a RDBMS.
> In theory, I thought it would work because it uses a JDBC driver. I plopped it into 3rdparty and setup the config (see bottom) and it saves without error. However it doesn’t show up in ‘show schemas;’ All attempts at querying it or showing tables fail.
> Anyone else crazy enough to attempt this?
> If you were wondering, I am interested in a connector that Presto has that Drill doesn’t, but Drill has better support for other things so I’d like to keep it on top.
>
> {
> "type": "jdbc",
> "driver": "io.prestosql.jdbc.PrestoDriver",
> "url": "jdbc:presto://presto-url:9999/catalog/schema <presto://presto-url:9999/catalog/schema>",
> "username": "user",
> "password": null,
> "caseInsensitiveTableNames": false,
> "sourceParameters": {},
> "enabled": true
> }