Try it free

Best practices for upgrading SLOs

  • Latest Dynatrace
  • Best practices
  • Published Aug 24, 2026

Classic SLOs use precomputed metrics tied to classic entity selectors. Rebuilt SLOs in Latest Dynatrace are backed by Grail, enabling advanced filtering, on-the-fly aggregation, and support for metrics not limited to precomputed values. These best practices help you extract the classic SLO expression, rebuild it in Latest Dynatrace, and validate parity before decommissioning the original.

Extract the SLO expression from the classic app

Before rebuilding an SLO in Latest Dynatrace, extract its metric expression from the classic Service-level objectives app.

  1. Open the classic Service-level objectives app.
  2. Select the SLO you want to upgrade.
  3. Open the metric in the metric explorer: this shows the full metric query expression the SLO uses.
  4. Click Open with and add the metric to a Notebook so you can inspect and copy the full expression.

The exported metric expression contains all the dimensions, filters, and aggregation logic needed to recreate the SLO in Latest Dynatrace.

Replace expression with sli when rebuilding

When you paste the copied metric expression into the Latest Dynatrace SLO app, replace the field name expression with sli. This is the only structural change required to adapt the metric query from the classic format to the Latest Dynatrace SLO format.

To rebuild the SLO

  1. Open the Latest Dynatrace Service-level objectives app.
  2. Click + Service-level objective and select + Custom SLO.
  3. Paste the metric expression copied from the classic app.
  4. Change expression to sli in the query.
  5. Click Refresh to preview the SLO calculation.
  6. Save the SLO.

The rebuilt SLO benefits from Grail-backed metrics, which support advanced filtering and aggregation not available in the classic SLO app.

After saving the rebuilt SLO, the team gains access to additional capabilities: Grail-backed metrics, advanced filtering, on-the-fly aggregation, and SLO definitions not limited to precomputed metrics.

Validate against classic baselines before decommissioning

Run both the classic and Latest Dynatrace SLOs in parallel for at least one full evaluation window before decommissioning the classic version.

To validate parity

  1. Compare SLO burn rates and metric values between classic and Latest Dynatrace versions over the same time range.
  2. Document any variance and agree on an acceptable threshold; small differences may result from how Grail-backed metrics calculate aggregations compared to classic precomputed metrics.
  3. Have the owning team confirm that the Latest Dynatrace SLO meets their operational requirements.
  4. Update any dashboard tiles or alert conditions that reference classic SLO metrics before decommissioning.

Only decommission the classic SLO after the team confirms parity and all dependent configurations have been updated to reference the Latest Dynatrace equivalent.

Note the Smartscape 2.0 dependency for entity references

The upgraded SLO metric query references classic entities. While the SLO is now running in Latest Dynatrace and backed by Grail, the entity selectors in the query still point to classic entity IDs.

A second update is required as part of the Smartscape 2.0 entity migration stage to replace classic entity references with Smartscape 2.0 entity IDs. Until that update is applied, the SLO continues to function correctly but depends on classic entity data remaining available.

Plan the Smartscape 2.0 entity reference update as a follow-up task when executing the remaining entities and metrics upgrade stage.

Ready for more?

Return to Upgrade configurations & settings to continue your upgrade.

Related tags
Dynatrace Platform