Streaming mode for thin JDBC driver is expected in Apache Ignite 2.5.
Meanwhile you can load data using thick driver which support streaming, and
then switch to thin driver for normal operations it you prefer it over
thick one.
On Fri, Dec 15, 2017 at 9:09 PM, Denis Magda wrote:
> Hi Michael,
>
Anyway, until the thin client is optimized I would suggest you to use the
client-based JDBC driver with “streaming” setting on for the preloading needs:
https://apacheignite-sql.readme.io/docs/jdbc-client-driver
—
Denis
> On Dec 15, 2017, at 10:09 AM, Denis Magda wrote:
>
> Hi Michael,
>
> Y
Hi Michael,
Yes, I heard from Ignite SQL experts that Ignite thin client is not optimal for
data loading yet. However, there is already some work in progress to speed up
the loading and batching of a myriad of rows.
Vladimir, could you please step in and comment here.
—
Denis
> On Dec 14, 20
With one server you just insert data locally. After you added the second
server you started moving data over the network. Apparently putting data
into local RAM is much faster than delivering same date to a remote node.
Ignite is first of all a distributed system. I suggest this: configure your
da
Hello, I am a new Ignite leaner. I want to insert 50,000,000 rows into a
table. Here,i got a problem.
When one host and one sever node, the speed of insert is about 2,000,000 per
minute, the usage of cpu is 30-40%; however two hosts and two sever nodes,
about 100,000 per minute,and the usage of cp
