> 2017-10-18 19:21 GMT+02:00 Donald Szeto :
>
>> Looks like an out-of-memory issue here. How much memory does the build
>> environment has?
>>
>
> A virtual server with 1GB in DigitalOcean . A system with 2GB of RAM would
> be enough?
>
>
>
Scala.
> I have KEYS file and source. Java version is "1.8.0_151".
>
>
> luciano@localhost:~/predic$ gpg --import KEYS
> gpg: key D3541808: "Suneel Marthi (CODE SIGNING KEY) "
> not changed
> gpg: key 8BF4ABEB: "Donald Szeto (CODE SIGNING KEY) &quo
Chiming in a bit. Looking at the serialization error, it looks like we are
just one little step away from getting this to work.
Noelia, what does your synthesized data look like? All data that is
processed by Spark needs to be serializable. At some point, a
non-serializable vector object showing i
Hey Noelia,
Since you are using PIO 0.11 you may want to use the previous template
version that was targeted for PIO 0.11. The latest version is targeted to
0.12’s default stack, which is Scala 2.11.
The only difference between these two template versions is targeting
different PIO versions. Ther
Dear PredictionIO community,
With your support and contribution over the past year, the PredictionIO
project is now very close to graduating to become a top-level project. The
Apache Incubator PMC has voted positively on our proposed graduation, and
it will be included in the ASF Board Meeting on
t Ferrel, did your team do something similar, use ids to go get
>> aggregate properties in the predict method? I prototyped with the universal
>> Recommendation template a couple months back and remember the predict
>> endpoint only required ids.
>>
>> Thanks for the supp
Hi Shane,
If you are not looking to do aggregative on Spark when you retrieve
additional information from event store, you probably should look at using
LEventStore that does not go through Spark. Depending on your use case, the
roundtrip time of involving Spark in your predict method might not be
Hi Noelia,
The URL should be http://localhost:7070/events.json?accessKey=myaccesskey
instead.
Regards,
Donald
On Tue, Oct 3, 2017 at 6:10 AM, Noelia Osés Fernández
wrote:
> Hi all,
>
> I'm new with Prediction IO and in order to understand how it works I'm
> customizing the Universal Recommende
Hi Shane,
Thanks for your finding! Perhaps we should improve the documentation or
even privatize the `fields` member.
The use of `get` is noted here:
https://github.com/apache/incubator-predictionio/blob/develop/data/src/main/scala/org/apache/predictionio/data/storage/PropertyMap.scala#L29
Regar
Hey,
I think you are probably looking for a passthrough argument like this.
pio deploy ... -- --conf "spark.driver.extraJavaOptions=-javaagent:..."
Please take a look at http://spark.apache.org/docs/latest/configuration.html
for more details.
Regards,
Donald
On Thu, Sep 28, 2017 at 11:31 AM, R
Congrats to everyone who is part of this, and Chan for his first Apache
release!
On Wed, Sep 27, 2017 at 2:40 PM Chan Lee wrote:
> The Apache PredictionIO team would like to announce the release of Apache
> PredictionIO 0.12.0-incubating.
>
> Release notes are here:
>
> https://github.com/apache
kawakibi
wrote:
> no.. i didnt make any changes ...
>
> On Thu, Sep 14, 2017 at 10:49 PM, Donald Szeto wrote:
>
>> Did you make any change to the template or using a stock one? Looks like
>> a Lucene incompatibility.
>>
>> On Thu, Sep 14, 2017 at 1:40
9200
> > #
> > # For more information, consult the network module documentation.
> > #
> > # ----- Discovery
> > --
> > #
> > # Pass an initial list of hosts to perform discovery when new node is
Hi Noelia,
Sorry that you ran into the orphaned Vagrant page. We haven’t had a chance
to update Vagrant yet, so we removed all links going to that page. If you
recall how you arrive at that page, please let us know.
Meanwhile, please install from source code, or try one of the
community-built Doc
Hey Jim,
Did you build PIO 0.12 with ES 1.4 support? ES 1.x is being deprecated in
0.12 so the default build will use ES 5.x support.
See the upcoming release notes:
https://github.com/apache/incubator-predictionio/blob/release/0.12.0/RELEASE.md#behavior-changes
Regards,
Donald
On Wed, Sep 20,
kSubmit$.org$apache$spark$deploy$SparkSubmit$$runMain(SparkSubmit.scala:731)
> at org.apache.spark.deploy.SparkSubmit$.doRunMain$1(SparkSubmit.scala:181)
> at org.apache.spark.deploy.SparkSubmit$.submit(SparkSubmit.scala:206)
> at org.apache.spark.deploy.SparkSubmit$.main(SparkSubmit.scala:121)
> at org.apache.spark.dep
Hi,
What version of Elasticsearch are you running? Do you see more information
from pio.log at where you ran "pio train"?
Regards,
Donald
On Thu, Sep 7, 2017 at 1:54 AM tajrij kawakibi
wrote:
> Hello,
>
> i just finish this 2 tutorial successfully,
>
>
> https://predictionio.incubator.apache.o
Hi Abhimanyu,
Have you try giving the JVM a larger heap memory, e.g. pio deploy --
--driver-memory 4g
Regards,
Donald
On Tue, Aug 29, 2017 at 9:06 AM, Abhimanyu Nagrath <
abhimanyunagr...@gmail.com> wrote:
> Hi ,
> I am new to predictionIo I am using template https://github.com/
> alexice/templ
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Mars Hall to become a PMC member and committer, and we are
pleased to announce that he has accepted.
Mars has been working on PredictionIO since 0.10 and has suggested and made
changes to the core so tha
+1. If some of them are found useful, they should be moved to examples/ and
have documentation about it.
On Thu, Jul 13, 2017 at 6:55 PM takako shimamoto
wrote:
> Hi,
>
> Examples under examples/experimental have been neglected for a long time.
> Probably no one won't make efforts for maintenanc
Many good discussions. Let me provide my input on these issues.
Multiple installations of PredictionIO should use different database names.
An analogy would be Wordpress installations that expect its own metadata
database. I understand the downside to this is that some users only have
access to on
Hi Thomas,
Supporting Kafka is definitely interesting and desirable. Are you looking
to sinking your Kafka messages to event store for batch processing, or
stream processing directly from Kafka? The latter would require more work
because Apache PIO does not yet support streaming properly.
Folks f
Hey Mars,
Thanks for the suggestion and I agree with your point on the metadata part.
Essentially I think the app and channel concept should be instead logically
grouped together with event, not metadata.
I think in some advanced use cases, event storage should not even be a hard
requirement as e
am assuming
> this is done in Serving.scala?
>
>
> On Fri, Jun 16, 2017 at 2:18 PM Donald Szeto wrote:
>
>> It also depends on how much you are willing to denormalize. If your
>> items' metadata don't change, it would be fine to send items' metadata in
>
access to JIRA?
>
> Best,
>
> Shane
>
> *Shane Johnson | 801.360.3350*
> LinkedIn <https://www.linkedin.com/in/shanewjohnson> | Facebook
> <https://www.facebook.com/shane.johnson.71653>
>
> On Fri, Jun 16, 2017 at 10:10 AM, Donald Szeto wrote:
>
>> Hi Shane
It also depends on how much you are willing to denormalize. If your items'
metadata don't change, it would be fine to send items' metadata in as part
of the event.
If you cannot denormalize and your scale is huge, you might want to
consider putting metadata in a KV store.
On Fri, Jun 16, 2017 at
Hi Shane,
Look for the query class that is part of the template (it is a type
parameter for the algorithm class). There you can customize your input
query to accept an array of objects, then modify the predict method in the
algorithm class to handle them.
I believe this is getting asked more that
Hi Dan,
Access key generation involves an entropy source which usually is pretty
small on VMs or containers. When entropy runs out, the random number
generator will block wait until there is entropy.
Try using http://www.issihosts.com/haveged/ to seed your entropy.
Regards,
Donald
On Sat, Jun 1
Unfortunately with OSS it is not easy to survey accurately the user base
unless we provide some sort of opt-in feedback (automatically or by putting
up a poll on the web site). We have many users who don't subscribe to
mailing lists.
I personally prefer updating source build defaults to a reasonab
Hi Sudhanshu,
What versions of Spark, Sparkling Water, PredictionIO are you running on?
The community might not have tested the template on the latest versions of
these technologies.
Regards,
Donald
On Tue, Jun 6, 2017 at 9:43 PM SUDHANSHU KAUSHIK <
sudhanshukaushik1.61...@gmail.com> wrote:
> H
Hey Emilio,
How are you storing your model? Failing to invert means it cannot
deserialize your model from disk. It's definitely an error message that we
could improve.
Regards,
Donald
On Tue, Jun 6, 2017 at 5:48 AM Emilio Dettori wrote:
> I'm working on a modified version of Text Classificatio
Hey all, this has a huge impact to the default build, so If you see any
issue with this, please let us know as soon as possible.
On Sun, Jun 4, 2017 at 10:25 PM, Shinsuke Sugaya
wrote:
> Hi all,
>
> We have a plan to change default build targets in PIO-83 and PIO-84.
> Current versions look too
Hey Andrew, is it possible for us to create non-release binary
distributions and ask Incubator PMC to help take a look without starting a
voting thread? This would allow us iron out issues in parallel to working
on 0.12.
On Wed, May 31, 2017 at 9:08 AM, Andrew Purtell wrote:
> Really glad to see
Hi Marius,
Test data points come from the `readEval` method.
References:
https://predictionio.incubator.apache.org/api/current/#org.apache.predictionio.controller.PDataSource
https://github.com/apache/incubator-predictionio-template-recommender/blob/develop/src/main/scala/DataSource.scala#L66
Re
Was this line correctly set?
https://github.com/apache/incubator-predictionio/blob/develop/conf/pio-env.sh.template#L30
On Sat, May 27, 2017 at 12:00 AM Mars Hall wrote:
> Also, re: the Postgres connection string, it is correct for my setup. My
> docker-compose creates a Postgres host named "po
Hey Shane,
One quick workaround is to manually edit this line for now:
https://github.com/apache/incubator-predictionio/blob/develop/bin/pio-shell#L62
and adding the JDBC assembly JAR after the main assembly JAR.
Sorry for the brief reply as I'm traveling. I will follow up with more
details whe
Thanks Pat for starting this. Added some edits to it. Everyone please feel
free to comment. PMCs please feel free to request edit access.
Very happy that we started the discussion!
On Thu, May 18, 2017 at 1:50 PM Pat Ferrel wrote:
> Should we apply for graduation to Apache TLP soon, like maybe
It is not clear to me from the gitbox page that we can move our source out
of ASF controlled servers and still be covered by ASF Legal. It's a pretty
big change if true. Did I miss an announcement from ASF?
Personally am happy to have things on GitHub because the flow would be a
lot simpler.
On T
up.
> >
> > He will talk about PredictionIO at JJUG Cross Community Conference
> > that is a largest Java technical conference in Japan this Saturday.
> > I'm looking forward to his talk!
> >
> > 2017-05-18 0:53 GMT+09:00 Donald Szeto :
> > > Correction
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Sara Asher to become a product manager, and we are pleased to
announce that she has accepted.
Sara is a Director of Product Management for Salesforce Einstein, where she
creates products that let people
Correction: Takahiro is not a committer of Apache Portals. Sorry for the
misinformation.
On Wed, May 17, 2017 at 8:50 AM, Donald Szeto wrote:
> Hi all,
>
> The Project Management Committee (PMC) for Apache PredictionIO
> (incubating) has asked Takahiro Hagino to become a PM
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Takahiro Hagino to become a PMC member and committer, and we are
pleased to announce that he has accepted. He is also a committer of the
Apache Portals project.
He has made major contributions to the Pre
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Takako Shimamoto to become a PMC member and committer, and we are
pleased to announce that she has accepted.
She has made major contributions to the PredictionIO 0.11.0 release by
adding Elasticsearch 5
You might want to try Docker on Windows with one of these community
contributed Docker container images:
http://predictionio.incubator.apache.org/community/projects/#docker-installation-for-predictionio
On Tue, May 16, 2017 at 12:07 PM Naresh Peddy wrote:
> Hi Team,
>
> Currently am using window
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Shinsuke Sugaya to become a PMC member and committer, and we are
pleased to announce that he has accepted. He is also a committer of the
Apache Portals project.
He has made major contributions to the Pre
We are going to publish a roadmap soon and one of roadmap items is to
support Spark ML pipelines natively in parallel of DASE.
If you are interested in helping developing the core code to support it,
please reach out in the dev mailing list.
On Wed, May 10, 2017 at 1:02 PM Ravi Kiran wrote:
> C
Did you build your source according to
http://predictionio.incubator.apache.org/install/install-sourcecode/ ?
On Mon, May 8, 2017 at 7:51 PM Some idoit on the internet <
branda...@gmail.com> wrote:
> I got it when I did ./bin/pio-start-all
> how do I fix it?
>
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Naoki Takezoe to become a PMC member and committer, and we are
pleased to announce that he has accepted.
He has made major contributions to the PredictionIO 0.11.0 release by
adding Elasticsearch 5 suppo
Once you've done make-distribution.sh there should be a new tarball named
PredictionIO-0.11.0-incubating.tar.gz at the root level of your codetree.
We recommend extracting and using that distribution instead of directly
from the codetree.
The behavior you are seeing is a bug anyway, and is already
Hi,
Can you try changing scalaVersion in the template's build.sbt?
Regards,
Donald
On Mon, May 1, 2017 at 5:17 PM Barthe van Doorn
wrote:
> PredictionIO-0.11.0-incubating
> scala.version=2.11.6
> spark.version=2.1.0
> elasticsearch.version=2.3.3
>
> When I try to build an engine, I get error [
Hi,
Sorry for the broken link. Please help file a JIRA issue if possible.
Meanwhile you can generate RDoc directly from Ruby SDK source code.
Regards,
Donald
On Fri, Apr 28, 2017 at 2:29 PM André Hatlo wrote:
> Hello
> I have a bachelor theisis with PIO, Spark and ROR.
> I'm trying to figure
Hi Pankil,
Looks like HDInsight runs Spark on YARN. You would need to point PIO to the
right Hadoop config files, and specify the correct master URL (yarn) when
you run "pio train".
Regards,
Donald
On Fri, Apr 28, 2017 at 12:12 PM Pankil Doshi wrote:
> Hi,
>
> Has anyone tried running predicti
The Apache PredictionIO team would like to announce the release of Apache
PredictionIO 0.11.0-incubating.
Release notes are here:
https://github.com/apache/incubator-predictionio/blob/v0.11.0-incubating/RELEASE.md
Apache PredictionIO (incubating) is an open source Machine Learning Server
built on
I haven't looked at the source code of the template but a quick look at the
output of the sanity check it said "total observation, 0". Do you have any
training data?
On Tue, Apr 11, 2017 at 4:14 AM YOUSSEF ASSATA
wrote:
> please why i'm getting this error
>
Is the java command on your search path pointing to your Java 8
installation?
In any case, you can set your JAVA_HOME environment variable to make sure
pio is using the right Java installation.
On Tue, Apr 11, 2017 at 12:21 AM Ravi Kiran wrote:
> I am starting new app using query
>
> pio ap
Send an email to user-unsubscr...@predictionio.incubator.apache.org.
On Mon, Apr 10, 2017 at 8:13 PM Pengcheng Fang wrote:
> how to unsubscribe ?
>
>
>
>
You can also create access keys for existing apps that have write
permissions to certain event names only. It is useful for client side event
collection, and is how some major analytics vendor JS SDKs limit client
side keys from polluting your event log.
Please take a look at `pio help accesskey`
Are you using the appName in engine.json that corresponds to the appId you
used when you did `pio import`? Setting a new appName in engine.json will
not automatically create a new app. Apps are created when you do `pio app
new`.
On Tue, Apr 4, 2017 at 8:13 PM Vaghawan Ojha wrote:
Hi,
Yes, I cou
setup through Ambari.
>
>
> On Mar 31, 2017, at 6:29 AM, Donald Szeto wrote:
>
> Hi Malay,
>
> Is your Spark cluster a standalone deployment or based on YARN?
>
> Regards,
> Donald
>
> On Thu, Mar 30, 2017 at 11:48 PM Malay Tripathi
> wrote:
>
>> Hello,
Hi,
A statistics of your event data's distribution would be helpful.
Regards,
Donald
On Tue, Mar 28, 2017 at 11:27 AM Vaghawan Ojha
wrote:
> Hi,
>
> I've added more than 30 events in the engine, and I'm using this
> template (
> https://github.com/apache/incubator-predictionio-template-rec
Hi Malay,
Is your Spark cluster a standalone deployment or based on YARN?
Regards,
Donald
On Thu, Mar 30, 2017 at 11:48 PM Malay Tripathi
wrote:
> Hello,
>
> I am running pio train *on an edge node* of distributed 8 node spark
> cluster & 3 node Hbase.
> When I run "pio train" the job runs but
Hi Amy,
Since event server keeps adding events to the backend, the storage will
grow indefinitely unless you implement some sort of data retention policy
that periodically.
In 0.11, there are two options for this situation:
- You may use SelfCleaningDataSource. Backing up your existing data is
hi
Hi Clifford,
To use a remote Spark cluster, use passthrough command line arguments on
the CLI, e.g.
pio train -- --master spark://your_master_url
Anything after a lone -- will be passed to spark-submit verbatim. For more
information try "pio help".
To use a remote Elasticsearch cluster, please
at gives out a classnotfound
> error for slf4jlogger,
>
> Best,
> -ZZ
> On Tue, Jan 24, 2017 at 2:11 PM Donald Szeto wrote:
>
> Hi,
>
> This question is more suitable for the dev list, so I am moving the thread
> there.
>
> You will need to use make-distribution.sh f
Hi,
This question is more suitable for the dev list, so I am moving the thread
there.
You will need to use make-distribution.sh for the command line to take
effect. The command line uses the assembly JAR produced by the
make-distribution.sh script.
It would be great to add some kind of support i
Do you have more detail logs from Spark executors?
Regards,
Donald
On Wed, Dec 14, 2016 at 3:26 AM Bansari Shah
wrote:
> Hi all,
>
> I am trying to use spark environment in predict function of ML engine for
> text analysis. It extends P2LAlgorithm algorithm. System works on
> standalone cluster
, Jan 19, 2017 at 9:16 AM Venkatesh Kumar Srinivasan <
venkat...@studiodiseno.com> wrote:
> Yes. I am calling from different script.
>
>
>
> Regards
>
> Venki
>
>
>
>
>
> *From: *Donald Szeto
> *Reply-To: *
> *Date: *Thursday, 19 January 2017 a
How are you calling CreateServer? Are you bypassing the command line
interface?
Regards,
Donald
On Thu, Jan 19, 2017 at 5:52 AM Venkatesh Kumar Srinivasan <
venkat...@studiodiseno.com> wrote:
> Hello
>
>
>
> I am trying to deploy a model by calling CreateServer class, but when
> class loader ini
& Regards,
> Yasothai Rasappan
> SSE *|* *yasoth...@vishwak.com*
> *Vishwak Solutions Pvt Ltd*
> ( *044 - **4222 2300* *| Extn - 439*
> *From:* Pat Ferrel [mailto:p...@occamsmachete.com ]
> *Sent:* 11 January 2017 00:24
> *To:* user@predictionio.incubator.apache.org
this implementation?? I really appreciate any
> comments about that.
>
> Do you have any reference or documentation to try Increase Spark driver
> memory locally??
>
> Best Regards,
>
> On Mon, Oct 31, 2016 at 12:51 PM, Donald Szeto wrote:
>
>> Hi Amado,
>>
>
31, 2016 at 6:41 PM, Donald Szeto wrote:
>
>> Hi Sebastiaan,
>>
>> Would this be sufficient for your use case: http://stackoverflow.com
>> /questions/26500539/scala-getting-field-and-type-of-field-of-a-case-class
>>
>> This can be put in a trait as a method and
Hi Sebastiaan,
Would this be sufficient for your use case:
http://stackoverflow.com/questions/26500539/scala-getting-field-and-type-of-field-of-a-case-class
This can be put in a trait as a method and mix in to different case
classes. If the name of the type signature is enough for your GUI that
w
Hi Amado,
You have several options:
1. Train on a real Spark cluster with more executor memory.
2. Increase Spark driver memory if you want to test locally.
3. Reduce number of iterations in engine.json.
Regards,
Donald
On Sun, Oct 30, 2016 at 7:42 AM, Amado Espinosa
wrote:
>
> Hi, I stuck tr
Hi Amal,
Unfortunately not all templates come baked in with evaluation. Try start
building your own by following
http://predictionio.incubator.apache.org/evaluation/. You can also
reference other template's evaluation, such as
https://github.com/apache/incubator-predictionio-template-recommender/b
Hi Amal,
If you have a lot of data to import in batch, you might want to check this
out: http://predictionio.incubator.apache.org/datacollection/batchimport/
The Python script goes through the Event Server, which uses the Storage
API. The HBase driver code is here:
https://github.com/apache/incub
Hi Georg,
This plugin will probably be deprecated for 0.10.0 and onward so that the
engine build can be easily built be all other IDEs. The old version plugin
will still work for now though.
Regards,
Donald
On Thursday, October 20, 2016, Georg Heiler
wrote:
> Hi,
>
> will the sbt plugin be upd
A binary release can be released once all transitive dependencies have gone
through license compatibility check and appropriate license files and
notices included. This is a lengthy and manual process so would
highly appreciate any help.
Regards,
Donald
On Wednesday, October 19, 2016, Pat Ferrel
Hi Amal,
For yarn-cluster mode, please use the --scratch-ufo argument and point it
to a location on HDFS where you have write access to. PIO will copy the
necessary files to the HDFS location for your yarn-cluster Spark driver to
read.
Regards,
Donald
On Tuesday, October 18, 2016, amal kumar wr
Hi,
This might be related to the misnamed parent artifact. Have you try again
and see if this has been resolved now?
Regards,
Donald
On Monday, October 17, 2016, infoquest india
wrote:
> I am also facing same issue. Please help.
>
> Thanks
> Gaurav
> http://www.infoquestsolutions.com
> Turning
Hi Georg,
Has this been resolved? It looks like a network glitch where SBT itself
cannot be downloaded.
Regarding installation, you will need to follow the official documentation
on building PredictionIO from source. A binary can be released when we have
all transitive dependencies go through lic
Hi Amal,
It looks like your PIO host cannot connect to your HBase host (
*ip-xxx-xx-xx-xx.us-west-2.compute.internal/xxx.xx.xx.xx:16020*). Could you
verify that the destination port is open and both hosts can communicate?
Regards,
Donald
On Thursday, October 13, 2016, amal kumar wrote:
> Hi,
>
Hi Bansari,
For production template, please used release artifacts on the central Maven
repository:
http://search.maven.org/#artifactdetails%7Corg.apache.predictionio%7Capache-predictionio-core_2.10%7C0.10.0-incubating%7Cjar
In your case, please use
"org.apache.predictionio"%% "apache-predic
Hi Bansari,
Looks like the underlying problem is ZooKeeper not being able to start. Are
you using the embedded ZooKeeper or a standalone ZooKeeper cluster? If you
could afford it, it might help to clear the ZooKeeper data and try again.
Regards,
Donald
On Mon, Oct 10, 2016 at 1:24 AM, Bansari Sh
The Apache PredictionIO team would like to announce the release of Apache
PredictionIO 0.10.0-incubating.
Apache PredictionIO (incubating) is an open source Machine Learning Server
built on top of state-of-the-art open source stack, that enables developers
to manage and deploy production-ready pre
Hi Amal,
Since you are using Redshift and their JDBC driver probably collides with
those provided from Postgres, you might have to use a different data
storage for your meta data and model data.
Regards,
Donald
On Wednesday, October 5, 2016, amal kumar wrote:
> Hello,
>
> I am trying to config
Hi Amal,
Could you try disabling IPv6 of your Java processes and see if it helps?
export JAVA_OPTS=-Djava.net.preferIPv4Stack=true
pio eventserver --ip 0.0.0.0
Regards,
Donald
On Thursday, October 6, 2016, amal kumar wrote:
> Hi Gustavo,
>
> I have tried both curl and web browser from outside
Hi Cristina,
If you are using 0.9.x, you can keep using the old "io.prediction" package
name for your engine source code. When you follow the tutorial online, just
use "io.prediction" in place of "org.apache.predictionio" wherever you come
across it.
If you are trying 0.10.0 though, make sure you
cluster. In this case does scratch-uri will work or we have to
> follow another process.
>
> Please suggest me.
>
> Thank you
> Regards,
> Bansari
>
> On Wed, Oct 5, 2016 at 1:30 AM, Donald Szeto wrote:
>
>> Hi Bansari,
>>
>> The --scratch-uri swit
AM, Bansari Shah wrote:
>
> Hi Donald,
>
> I am running my spark cluster of 3 node with YARN and spark driver is
> within cluster.
>
> Thanks
> Regards,
> Bansari
>
> On Tue, Oct 4, 2016 at 9:59 PM, Donald Szeto wrote:
>
>> Hi Bansari,
>>
>> How
Hi Bansari,
How are you running your Spark cluster? Standalone, YARN, or Mesos? Are you
running the Spark driver on the client or within the cluster?
Regards,
Donald
On Tue, Oct 4, 2016 at 5:55 AM, Bansari Shah
wrote:
> Hi,
> Can you please guide me how to use 'scratch-uri' argument in case of
Hi Jenna,
The AWS image is no longer supported as of now. You may set up a raw AWS
EC2 instance and install PredictionIO there.
Regards,
Donald
On Mon, Oct 3, 2016 at 2:36 PM, Jenna Kwon wrote:
> Hello everyone,
>
> I would love to deploy my engine and was wondering what the current state
> fo
Hi Jenna,
You probably want to have a reverse proxy in front of your app
(localhost:3000) so that it is exposed to the Internet via standard ports
(80 for HTTP or 443 for HTTPS). Your app can communicate to the event
server and engine over localhost. They don't need to be exposed to the
Internet.
Hi Cristina,
Sorry for the confusion. We are pretty close to our first official release.
We will send out a note and update all docs once that happens.
Regards,
Donald
On Mon, Oct 3, 2016 at 9:05 AM, Chan Lee wrote:
> Hi Cristina,
>
> You're right. The classpath change is part of v0.10 which i
Sorry for side-tracking. I think Kappa architecture is a promising
paradigm, but including batch processing from the canonical store to the
serving layer store should still be necessary. I believe this somewhat
hybrid Kappa-Lambda architecture would be generic enough to handle many use
cases. If th
Hi Hasan,
Does your randomForestModel contain any RDD?
If so, implement your algorithm by extending PAlgorithm, have your model
extend PersistentModel, and implement PersistentModelLoader to save and
load your model. You will be able to perform RDD operations within
predict() by using the model's
ing "No automatic version matching of PIO binary distribution and
>> artifacts version used in the engine template":
>>
>> The Heroku slug contains the PredictionIO binary distribution used to
>> build the engine, so there's never a version matching issue. I guess
Hi all,
I want to start the discussion of removing engine registration. How many
people actually take advantage of being able to run pio commands everywhere
outside of an engine template directory? This will be a nontrivial change
on the operational side so I want to gauge the potential impact to
Hi all,
The Project Management Committee (PMC) for Apache PredictionIO (incubating)
has asked Paul Li to become a PMC member and committer, and we are pleased
to announce that he has accepted.
Paul Li was a regular committer to the PredictionIO codebase before its
donation to the Apache Software
pio build and pio train are independent processes from pio eventserver and
pio deploy, so you can still take events and serve queries.
On Thu, Sep 8, 2016 at 11:53 AM, Gustavo Frederico <
gustavo.freder...@thinkwrap.com> wrote:
> Is the server available for events and queries while executing pio
Hi,
If you are using the stock classification engine template, your easiest way
out would be to use multiple engine servers. If that is not desired, you
could modify the engine template to handle different data sets
independently.
Regards,
Donald
On Tue, Sep 6, 2016 at 5:49 AM, Dipen Patel wrot
1 - 100 of 106 matches
Mail list logo
