# dbt snapshots: How to backfill from a table without updated\_at/created\_at timestamp?

**URL:** <https://discourse.getdbt.com/t/dbt-snapshots-how-to-backfill-from-a-table-without-updated-at-created-at-timestamp/9347>\
**Category:** Help\
**Tags:** snapshots\
**Created:** [July 30, 2023, 11:13pm UTC](https://discourse.getdbt.com/t/dbt-snapshots-how-to-backfill-from-a-table-without-updated-at-created-at-timestamp/9347 "2023-07-30T23:13:24Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![philippefutureboy](https://avatars.discourse-cdn.com/v4/letter/p/a9adbd/32.png) [@philippefutureboy](https://discourse.getdbt.com/u/philippefutureboy)\
**Post date:** [July 30, 2023, 11:13pm UTC](https://discourse.getdbt.com/t/dbt-snapshots-how-to-backfill-from-a-table-without-updated-at-created-at-timestamp/9347/1 "2023-07-30T23:13:24Z")

</div>

_Note: This issue has also been posted on StackOverflow, and can be found [here](https://stackoverflow.com/posts/76799911)._

We have several years worth of daily `SELECT * FROM source.table` captures for a given table. We want to create a dbt snapshot that tracks row level changes over the set of daily captures. The table does not have an updated\_at field.

The perceived issue is that:

1. Given that the table does not have an created\_at/updated\_at field, we cannot use the dbt snapshot “timestamp” strategy;
2. Given that we have to backfill several years of daily snapshot on a first run, we think that we cannot use the “check” strategy, as our understanding is that the check strategy uses the time of execution to fill the dbt\_\* fields.

We could, in theory add the extraction timestamp as the updated\_at field. However since our understanding of the “timestamp” strategy is that dbt only tracks the timestamp field, we assume that every record will have a new version for every capture we add to the dbt snapshot table. Is that a correct assessment?

Our best option at this time is to update manually the values in dbt\_\* fields to the extraction timestamps of our captures in storage in between every dbt snapshot run, and use the “check” strategy to determine changes.

Is there any dbt- native solution possible for this issue?

Thanks a lot for your answers!