Trigger : update column based on value of another column - sql

Question

We have a table that contains a status column, and associated to that is a column to track when the status changed values.
As an example, we have a status of 'Off' and a status of 'Found Off' along with associated columns, DateOff and DateFoundOff. I'm trying to create a trigger to update these date columns when the status changes.
It seems rather straightforward to me, but what is occurring is when the Status changes the associated date column updates correctly, but the other date column becomes null. So if I change Status = 'Off' DateOff has the correct date but DateFoundOff becomes null and visa versa.
I created two triggers - first one is:
ALTER TRIGGER [GIS].[UPDATE_FOUNDOFF]
ON [GIS].[METEROUTAGEPOINTS]
AFTER UPDATE
AS
IF (UPDATE (OutageStatus))
BEGIN
SET NOCOUNT ON;
UPDATE [gis].[METEROUTAGEPOINTS]
SET DateFoundOff = CURRENT_TIMESTAMP
FROM gis.METEROUTAGEPOINTS mop
INNER JOIN inserted AS i ON i.ConObject = mop.ConObject
WHERE i.OutageStatus = 'Found Off'
END
And the second
ALTER TRIGGER [GIS].[UPDATE_DATES]
ON [GIS].[METEROUTAGEPOINTS]
AFTER UPDATE
AS
IF (UPDATE (OutageStatus))
BEGIN
SET NOCOUNT ON;
UPDATE [gis].[METEROUTAGEPOINTS]
SET DateOff = CURRENT_TIMESTAMP
FROM gis.METEROUTAGEPOINTS mop
INNER JOIN inserted AS i ON i.ConObject = mop.ConObject
WHERE i.OutageStatus = 'Off'
END
I simply do not understand how one trigger is changing the value of the Date column to null that is not associated to the current status value.
Thanks.
Edit: The issue was found to not be with trigger but instead be with how the tool being used to edit the data was holding onto something. Not sure I understand why, but by changing the edit workflow the problem was resolved. Marked answer as correct based on it giving a far better way to write the trigger

Answer 1

Neither of your update statements change DateFoundOff or DateOff to null. Something else must be going on.
However I would improve your trigger as follows:
Only use a single trigger, every trigger has overhead, having one trigger with one update statement will run faster than 2.
You aren't actually checking that the status has changed, all you are checking is that the update included that column. Add a check against the deleted table actually checks whether the value changed.
ALTER TRIGGER [GIS].[UPDATE_FOUNDOFF]
ON [GIS].[METEROUTAGEPOINTS]
AFTER UPDATE
AS
BEGIN
-- Avoid doing any processing if no rows are updated
IF NOT EXISTS (SELECT 1 FROM Inserted) RETURN;
SET NOCOUNT ON;
IF UPDATE(OutageStatus) BEGIN
UPDATE [gis].[METEROUTAGEPOINTS] SET
DateFoundOff = CASE WHEN i.OutageStatus = 'Found Off' AND d.OutageStatus <> 'Found Off' THEN CURRENT_TIMESTAMP ELSE DateFoundOff END
, DateOff = CASE WHEN i.OutageStatus = 'Off' AND d.OutageStatus <> 'Off' THEN CURRENT_TIMESTAMP ELSE DateOff END
FROM gis.METEROUTAGEPOINTS mop
INNER JOIN inserted AS i ON i.ConObject = mop.ConObject
INNER JOIN deleted AS d ON d.ConObject = mop.ConObject;
END;
END;