I've used Hive in a multiple connections per server instance setup. It
works ok, but it is a little flakey. I have some snapshot of trunk >
0.8.0 deployed. When I have some time, I'd like to help increase the
test coverage for multithreaded clients.
On 28 March 2012 19:19, Abhishek Pratap Singh <[EMAIL PROTECTED]> wrote:
> AFAIK hive does not support concurrency however Multiple connection can be
> made to HiveServer.
> Hive or i should say Hadoop is write once and read multiple architecture.
> Hive does not have a build in support for DELETE OR UPDATE.
> The inherent architecture of these kind of solution works best when you
> don't need Concurrency, Transaction or ACID properties. I have seen usecases
> where wrapper to handle transaction is implemented but at the significant
> cost of Performance.
> On Wed, Mar 28, 2012 at 11:03 AM, sumadhur <[EMAIL PROTECTED]> wrote:
>> Resending due to lack of response.
>> From: sumadhur <[EMAIL PROTECTED]>
>> To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
>> Sent: Tuesday, March 27, 2012 2:14 AM
>> Subject: Hive server concurrency question
>> Does Hive server support multiple concurrent client connections?
>> The following page says it does not support them.
>> Hive Server is currently not thread safe
>> The following page says that it was designed to support multiple
>> connections. But there's no test coverage.
>> HiveServer is multithreaded and was designed to support multiple
>> concurrent client connections. However, there is currently no test coverage
>> for this scenario. If you want to play it safe we recommend running one
>> HiveServer process per client connection.
>> The following JIRA seems to imply that the hive server does support
>> multiple concurrent connections.
>> What's the official stance on this? supported or not supported?
>> Are there any known issues with hive server's handling of multiple
>> concurrent client connections?