Data ReViewer Service can only run one config???

[RandyCrumley1]

So I have many users direct connecting to many different cells of data and they are updating the data.

If I want the server to run the checks over night, I have to use the datareviewer service controller.  Makes sense.

BUT IT ONLY WRITES TO ONE SESSION AND ONLY ALLOWS FOR ONE JOB???

WHAT??  How is that any help whatsoever.  That is the exact same as just going to arcmap and using the datareviewer interface.  Am I just supposed to sit around and watch the service controller

If I build RBJ checks to check their work and want to use the service to run the checks I have to -

Create an SDE datareviewer

Create an RBJ for 'AOI 1'

Remap and save a copy of the RBJ for 'AOI2, AOI3, AOI87...'

Load EVERY SINGLE ONE OF THOSE RBJS into the Reviewer controller

Let the service run and shove all of the results in the same session.

Get here before everyone else and relink the qc grid since the checker doesn't auto populate that even though it's linked

have the user re link the grid too

Recreate a the reviewer database every now and then so it doesn't get an unusable amount of records in it.

Update 80 different RBJ checks every-time I need to add or change a check, since each rbj has to be mapped to which spacific database it's being run on.

And don't say use the product library.  I've set up 20 different product libraries with 100 different data models with 50 different permission levels and it all it does is crash the controller.  I can't even get past the testing phase.

Thanks for nothing.

[KumarBhardwaj]

Hi Randy,

Please confirm the following to better understand your workflow:

1. Version of ArcGIS Data Reviewer you are using.

2. What are you changing in the batch job when you "Remap and save a copy of the RBJ for 'AOI2, AOI3, AOI87...'". Are you changing the data workspace for the batch job?

3. Do you want the data for each cell written to a separate session? Do you also want different versions for each sessions?

Thanks,

Kumar

[RandyCrumley1]

Hi Kumar.

Although I have some of the ArcGIS for Server software installed, we have not yet begun the long and arduous journey to migrating to a webserver (which, as far as I understand it, is required to use workflow manager and other ArcGIS for server features), so for now, I am just working with SQLServer 2008r2 and ArcGIS version10.2.2 for desktop.  To answer your questions:

1-

ArcGIS 10.2.2 for Desktop

ArcGIS 10.2.2 License Manager

Esri Production Mapping 10.2.2

Esri Defense Mapping 10.2.2

ArcGIS 10.2.2 Data Reviewer for Desktop

ArcSDE 10.2.2 for Microsoft SQL Server

ArcGIS 10.2.2 for Server

ArcGIS 10.2.2 Workflow Manager for Server

ArcGIS 10.2.2 Data Reviewer for Server

2 - Yes.  I am just remapping the work-space location so that the same checks will be valid to run on multiple different databases. 

I realize this is not the case, but ideally I would like to:  1 - create one set of checks in a master RBJ file 2-  pick which databases the checks get ran on and 3 - just specify the session (or even a different database for that matter) that each database's results get put into. 

3 - Yes. At minimum, if each cell's results would be written to different sessions, then the analyst responsible for that cell will be able to just connect to the reviewer database (an sde database) and see the check results for their area only (since each session will be unique to each cell).  I have it set to only run on changed features, so every day, only new results are added to the table.

But as it sits, I can set one reviewer work-space, that writes to one session, that uses checks that have to specify what database they are to be run on.  Database.  Not Databases.  So, as a simple example. if I want to check make sure none of the buildings in my databases have an ARA = 12, then I have to write that in an RBJ file for database one.  Then I have to duplicate that RBJ and change the database it's mapped to to database2. And so on.  And that would be okay, but from there all I can do is import all of those checks into the configuration.  But now what?  Where does that leave me?

It leaves me in one of two places.

I can either load all the checks for all the databases and have the service run that once.  But that just shoves all of the data into one session that people have to sort through and is separated geographically by vasts swaths of earth and man-made boundary. 

OR

I can load the RBJ that's mapped to area1 only and set that to either a different datareviewer workspace or a different session.  doesn't matter.  Then I could create a configuration file with just the checks that are mapped to area2 and set that to either a different datareviewer workspace or a different session.  doesn't matter. .  BUT the service controller only allows one configuration file at a time.  So unless I only want it to run on AREA1 every day, I have to wait for it to run area1's configuration.  Then change the configuration to Area2's configuration. Manually restart the service.  Wait for it to run on area2.  Change the config to AreaX. restart service...and so on. 

Which, at that level of manual over-site, you'd just be better off running the process in ArcMap because it's as or more labor intensive.

I hope this has helped you gain a clearer picture of my issue.  Do not hesitate to ask for more information.

[RandyCrumley1]

Furthermore, since posting this (since I had to give up on doing it the other way).  I had the idea of creating the RBJs for each area and installing the service on each analysts computer and having them initialize the service based on provided rbj checks that I build and just having them either map it to a session or different db.  doesn't matter.  BUT, surprise.  If it's installed on a desktop, and not directly on the server, the checks all set up, but the service never runs the check.  just flat out doesn't do anything.  pops up says service started, but never executes the job.  and it's a super basic job.  takes the server like 12 seconds to run it.

So, really, I'm back to square one, since none of this seems to work.  Which square one is like in 2010 or something when we just had analysts sit and wait while it ran for hours.  Which is really depressing.  Okay, I have to go now before I start in about having the option to set merge policies, but not being able to actually use them unless you want to invent a merge tool...

[MichelleJohnson · Accepted solution]

Don't use the Reviewer Service (It's being deprecated anyway).  Create a GP model and export it out to a python executable and use Windows task scheduler to run the check after hours.  When you come in, the results will be ready for you.

With the GP model, you can create a new session for each batch job you execute.

[RandyCrumley1]

Thank you.  I know how picky DataReViewer is about always wanting you to use the DataReViewer interfaces so I was convinced I had to go through the service controller.  I will try the GP to Script method - it sounds right to me.  Good call on phasing out the service, it seems like it wants to do things in the wrong order anyway.

-Randy

[KumarBhardwaj]

Hi Randy,

Just adding to what Michelle explained, you can refer to the following tutorial (please make necessary changes, if required as this is for version 10.3):

Running batch jobs as a scheduled task

http://desktop.arcgis.com/en/desktop/latest/guide-books/extensions/data-reviewer/exercise-3-running-...

Also, if you are planning for Data Reviewer for Server, you can use Batch Validation Manager which is an application that is appropriate for your workflow. Please refer to the following link for more details:

Data Reviewer Batch Validation Manager

http://www.arcgis.com/home/item.html?id=867d06e3237d41c19fb590852a6bd3d3

Thanks,

Kumar