predictionio-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Vaghawan Ojha <>
Subject Re: Deploying Multiple App in Same Machine
Date Thu, 04 May 2017 07:47:31 GMT
Hi Pat,

Thanks, I had the same Index names in engine.json. I changed that now and
retrain them, appnames are different, so that shouldn't be the problem. But
the problem still persists.

It's quite weird.

On Mon, May 1, 2017 at 11:33 PM, Pat Ferrel <> wrote:

> make sure the 2 apps have separate appNames in `pio app list` and have
> different indexNames in their engine.jsons.
> 1) appNames are where to get input, if the same then only params will be
> making different models, not different data. A PIO “app” is really a
> dataset.
> 2) indexNames tell the UR where to store the model in Elasticsearch, if 2
> engin.jsons have the same indexName, the latest `pio train` will overwrite
> the previous one.
> On Apr 30, 2017, at 10:13 PM, Vaghawan Ojha <> wrote:
> Hi,
> I'm using two seperate Universal recommender app in same machine.
> Everything is fine, but whenever I retrain one app, the score in the other
> app always goes to zero. However recommendation seems fine as far as I
> could see.
> Why this happens? I'm using different port to deploy the different apps.
> Have anyone else gone through the same problem?
> --
> You received this message because you are subscribed to the Google Groups
> "actionml-user" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to
> To post to this group, send email to
> To view this discussion on the web visit
> msgid/actionml-user/CA%2B69RXY0ncgUngGai5ztBwsqSwt%
> <>
> .
> For more options, visit

View raw message