refactoring: minor refactoring of PR #8916

CHANGELOG.md

@@ -19,6 +19,7 @@
 * **Prometheus**: Add support for instant queries [#5765](https://github.com/grafana/grafana/issues/5765), thx [@mtanda](https://github.com/mtanda)
 * **Cloudwatch**: Add support for alerting using the cloudwatch datasource [#8050](https://github.com/grafana/grafana/pull/8050), thx [@mtanda](https://github.com/mtanda)
 * **Pagerduty**: Include triggering series in pagerduty notification [#8479](https://github.com/grafana/grafana/issues/8479), thx [@rickymoorhouse](https://github.com/rickymoorhouse)
+* **Timezone**: Time ranges like Today & Yesterday now work correctly when timezone setting is set to UTC [#8916](https://github.com/grafana/grafana/issues/8916), thx [@ctide](https://github.com/ctide)
 
 ## Minor
 * **SMTP**: Make it possible to set specific EHLO for smtp client. [#9319](https://github.com/grafana/grafana/issues/9319)
@@ -31,7 +32,7 @@
 
 # 4.5.2 (2017-09-22)
 
-## Fixes 
+## Fixes
 * **Graphite**: Fix for issues with jsonData & graphiteVersion null errors [#9258](https://github.com/grafana/grafana/issues/9258)
 * **Graphite**: Fix for Grafana internal metrics to Graphite sending NaN values [#9279](https://github.com/grafana/grafana/issues/9279)
 * **HTTP API**: Fix for HEAD method requests [#9307](https://github.com/grafana/grafana/issues/9307)
@@ -44,7 +45,7 @@
 * **MySQL**: Fixed issue with query editor not showing [#9247](https://github.com/grafana/grafana/issues/9247)
 
 ## Breaking changes
-* **Metrics**: The metric structure for internal metrics about Grafana published to graphite has changed. This might break dashboards for internal metrics. 
+* **Metrics**: The metric structure for internal metrics about Grafana published to graphite has changed. This might break dashboards for internal metrics.
 
 # 4.5.0 (2017-09-14)
 
@@ -73,7 +74,7 @@
 ### Breaking change
 
 * **InfluxDB/Elasticsearch**: The panel & data source option named "Group by time interval" is now named "Min time interval" and does now always define a lower limit for the auto group by time. Without having to use `>` prefix (that prefix still works). This should in theory have close to zero actual impact on existing dashboards. It does mean that if you used this setting to define a hard group by time interval of, say "1d", if you zoomed to a time range wide enough the time range could increase above the "1d" range as the setting is now always considered a lower limit.
-* **Elasticsearch**: Elasticsearch metric queries without date histogram now return table formated data making table panel much easier to use for this use case. Should not break/change existing dashboards with stock panels but external panel plugins can be affected. 
+* **Elasticsearch**: Elasticsearch metric queries without date histogram now return table formated data making table panel much easier to use for this use case. Should not break/change existing dashboards with stock panels but external panel plugins can be affected.
 
 ## Changes
 

public/app/features/dashboard/time_srv.ts

@@ -195,7 +195,8 @@ class TimeSrv {
       from: moment.isMoment(this.time.from) ? moment(this.time.from) : this.time.from,
       to: moment.isMoment(this.time.to) ? moment(this.time.to) : this.time.to,
     };
-    var timezone = this.dashboard && this.dashboard.getTimezone ? this.dashboard.getTimezone() : 'local';
+
+    var timezone = this.dashboard && this.dashboard.getTimezone();
 
     return {
       from: dateMath.parse(raw.from, false, timezone),

public/app/features/panel/metrics_panel_ctrl.ts

@@ -217,10 +217,8 @@ class MetricsPanelCtrl extends PanelCtrl {
       "__interval_ms":  {text: this.intervalMs, value: this.intervalMs},
     });
 
-    var timezone = this.dashboard.getTimezone ? this.dashboard.getTimezone() : 'local';
-
     var metricsQuery = {
-      timezone: timezone,
+      timezone: this.dashboard.getTimezone(),
       panelId: this.panel.id,
       range: this.range,
       rangeRaw: this.range.raw,