Deploying Multiple App in Same Machine

[Vaghawan Ojha]

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? 

[Pat Ferrel]

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.

--
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 actionml-use...@googlegroups.com.
To post to this group, send email to action...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/actionml-user/CA%2B69RXY0ncgUngGai5ztBwsqSwt%3DZTyUcj%2B0ccqhzi_bqS%3D01A%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

[Vaghawan Ojha]

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 <p...@occamsmachete.com> 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 <vagha...@gmail.com> 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 actionml-user+unsubscribe@googlegroups.com.

[Vaghawan Ojha]

Sorry, it did work. the content was coming from the cache. Thank you very much, I just changed the ES indexNames. 

Thank you