Native Connection to Hive with ER/Studio Data Architect

I've been able to connect ER/Studio to Hive via ODBC to reverse engineer objects, but some of the data types aren't coming through properly.

I've been told that the native connection works better, but I'm having trouble getting the right connection parameters setup in the Reverse Engineering Wizard.

Has anyone successfully connected ER/Studio Data Architect to Hive natively in order to reverse engineer?

  • We have a video that shows how this is done.

    If you are still having trouble with this, please contact Support for assistance. Thanks!
  • In reply to Joy Ruff:

    Hi Joy,

    Thanks for the video. Unfortunately, we're using Kerberos here and the video connects without the use of Kerberos. I had already opened a case with support (Case#00538542) but didn't get a lot of help via that channel. I think I'll just need to talk with the Hadoop engineers here to find the right settings.

  • In reply to tomf:

    Hi Tom,

    We are also having the issue of being unable to connect ER/Studio to Hive now that we've got Kerberos on our Hadoop cluster. Were you ever successful in connecting? If so, would you be able to share the secret? We had no trouble connecting before Kerberos was installed, but now it seems no matter what parameters we put in, nothing works. Thanks for any help or guidance.

  • In reply to mfite:

    Hi Michaela,

    What I understand from talking with Idera, only Hive 0.12 and 0.13 are supported by ER/Studio Data Architect for native connections. We're using a higher version of Hive which is not compatible. Earlier this year Ron Huizenga from Idera did mention that they plan to add more platform support in 2018.

  • In reply to Joy Ruff:

    In the video, it shows the map and array data types listed in the DDL from the reverse engineering step. But when I try to create a new table, I don't see these two data types listed. Shouldn't they be included in the list of supported data types?