Error reading completed NN results

[augurworks1]

NN completes, but getting error when clicking url to read results. restarted the alfred and tomcat. I vaguely recall this happening recently where the augtrain.augout file was empty... or something like that.

In checking the augtrain.augout and getting NULL on the first day and last day of the date range. Not sure if this means anything.

It's occurring on both Dev and Prod.

[TheConnMan]

The issue is that the results of the Neural Net are stored in files which live in the web-apps directory. When a new version gets deployed the whole app folder is removed and redeployed. It is manifesting because we're deploying much more frequently. I'll have to either store the files outside the application folder or store them in the database.

[augurworks1]

I haven't seen this for quite some time though. Why would the result list
still exist in the UI if the folder containing the files was removed?

On Sun, Mar 15, 2015 at 7:39 PM, Brian Conn notifications@github.com
wrote:

The issue is that the results of the Neural Net are stored in files which
live in the web-apps directory. When a new version gets deployed the whole
app folder is removed and redeployed. It is manifesting because we're
deploying much more frequently. I'll have to either store the files outside
the application folder or store them in the database.

—
Reply to this email directly or view it on GitHub
#202 (comment).

Drew Showers | AugurWorks CEO
609-304-1341 (m) | www.augurworks.net

[TheConnMan]

Yes, they would still be in the UI.

[augurworks1]

This would even cause error for new NN results, post deployment?

On Sun, Mar 15, 2015 at 7:59 PM, Brian Conn notifications@github.com
wrote:

Yes, they would still be in the UI.

—
Reply to this email directly or view it on GitHub
#202 (comment).

Drew Showers | AugurWorks CEO
609-304-1341 (m) | www.augurworks.net

[TheConnMan]

Fixed, but there is an issue with the Neural Nets getting removed after a deploy. I've added the issue at #203.