Any ideas how I might troubleshoot this: I have Ma...
# troubleshooting
i
Any ideas how I might troubleshoot this: I have Matatika's tap-googleads running, and it seems to work just fine when running
meltano invoke tap-googleads
I get records just fine. I have only selected
stream_campaign_performance.*
The records look like
Copy code
{
  "type": "RECORD",
  "stream": "stream_campaign_performance",
  "record": {
    "campaign": {
      "resourceName": "customers/12345678/campaigns/12345678913245",
      "status": "PAUSED",
      "name": "this is just the campaign name"
    },
    "metrics": {
      "clicks": "0",
      "costMicros": "0",
      "ctr": 0,
      "impressions": "10"
    },
    "segments": {
      "device": "DESKTOP",
      "date": "2021-09-10"
    }
  },
  "time_extracted": "2022-09-01T08:49:13.695805Z"
}
But when I do an elt to target-snowflake, I get an error:
Copy code
2022-09-01T08:53:16.981267Z [info     ] invalid identifier 'RESOURCE_NAMES' cmd_type=loader name=target-snowflake run_id=6d80b3c8-0e20-4cb6-b5fe-4946b294199d state_id=2022-09-01T085251--tap-googleads--target-snowflake stdio=stderr
Anyone have an idea where to look?
a
@daniel_walker any thoughts? @ilkka_peltola we setup and run this tap daily with Snowflake, but don't think we select any streams in our tests. Can you confirm you are using the pipelinewise snowflake target?
i
yes, pipelinewise
d
We run the tap against a different target-snowflake (a fork of the old Meltano target that doesnt require s3). But It looks like this is a snowflake case sensitive column name issue like: https://community.snowflake.com/s/question/0D50Z000090md2oSAA/sql-compilation-error-invalid-identifier. So I guess the pipelinewise target-postgres is creating your Snowflake columns in a different way.
i
I don't really mind going to a different snowflake connector, if it works. The one I'm currently running (pipelinewise variant) has worked with all other sources except for this one, tap-googleads. But so I guess I should be raising an issue with the pipelinewise guys then? What snowflake target variant is most robust for now?
d
Pipelinewise variant of target-snowflake is the most supported for sure. I've not run it before though and don't have anything s3 setup so can't really test what exactly is going wrong or what a fix would be on the tap side. What taps have you been using, just so I can compare to tap-googleads and see If I can spot a difference?
i
I'm running • tap-google-analytics (meltano) • tap-google-sheets (singer) • tap-mongodb (transferwise) • tap-stripe (meltano) • tap-zendesk (twilio-labs)
d
Hey I've taken a bit of a look into this and get the same error that you are getting. Im going to open some issues on our repo to get the tap working with the pipelinewise variant. For the time being it does work with our version of the snowflake loader if you want to get it work asap: https://github.com/Matatika/target-snowflake. Note this target has slightly different setting names, also its pretty basic.
a
Hi @daniel_walker I'm unfortunately encountering the same issue. Any SLA for getting the tap working with pipelinewise?
d
Hey Angus, sorry we currently don't have an eta on the fix for supporting Pipelinewise
target-snowflake
. It's on the to-do list for sure but we have a few big things going on at the moment. If you're willing though we'll accept a PR for the issue. Like I said above we only currently test it with our
target-snowflake
directly, so you could try using that? If not another quick solution might be to use @visch's
tap-googleads
that we forked to create ours: https://github.com/AutoIDM/tap-googleads.
a
Perfect will try that, thank you for the quick response