Connector Improvement: Add log event when connector code changes are deployed
Answered
It would be great to add log events when the code for a connector changes, including information about the connector code version.
Currently, there are no customer-visible logs that reflect when the Fivetran code for a connector changes.
We recently had major trouble with a connector that failed syncing due to a bug in Fivetran's connector code. It took us a week to identify this as the cause. Having visibility into when the connector code changes would have made it much easier for us to see that the failures closely followed the new connector version being rolled out.
It would be very useful to us to have visibility into what version of the connector code we are running, especially since the Fivetran team downgraded our connector to restore operation, but we don't have any way to see this directly.
Thanks!
-
Official comment
Hi Andy,
Thanks for writing in. This issue you've raised is one we take very seriously, as confidence and trust in our product is a key benefit of a managed data pipeline. Currently, we report behavior changes in our change log which can be filtered for specific connectors.
We are continuously evaluating our performance with regards to bugs, regressions, and QA. We aspire to provide a unified cloud SaaS experience in which versioning is irrelevant. That said, I understand where you're coming from and you can rest assured the Product Team is aware of the perspective you represent here. We'll keep this need in mind and continue to review to any other experiences that customers wish to share either in this thread, with their account managers, or elsewhere.
All the best,Ray Harris
Product Manager, Sales & Customer Success Analytics -
Thanks Ray
I don't see anything relevant in the change log for the Postgres connector, which is what bit us here. How complete are those change logs?
-
Hey Andy,
PM for the Postgres connector here. The change logs usually represent new feature rollouts.
Looking at your account, I see this issue was due to a new, refactored Postgres connector we have been automatically rolling out to customers throughout the past 3 months. I agree that we should share more information as we deploy code changes and modify connector functionality, so I will be sure to share this feedback with the team.
Please sign in to leave a comment.
Comments
3 comments