Showing results for 
Search instead for 
Did you mean: 
Level: Powered On

POSTGRES ODATA time comparison

I am trying to get all records from a postgres database where the date_completed (timestamp without timezone) is between startoftoday and endoftoday. However I cannot get the ODATA query for less than or greater than to work. I keep getting this error. 


  "status": 400,
  "message": "We cannot apply operator < to types DateTime and DateTimeZone.\r\n     inner exception: We cannot apply operator < to types DateTime and DateTimeZone.\r\nclientRequestId: 12f42943-0acf-4b4a-a482-3d72ca9dbf98",
  "source": ""

The odata filter is equal to 

utcnow('yyyy-MM-ddTHH:mm:ssZ’) lt date_completed

 This is the odata filter after the expressions run

2018-09-06T08:07:16Z lt date_completed
Community Support Team
Community Support Team

Re: POSTGRES ODATA time comparison

Hi @elksson,


Could you please share a screenshot of the configuration of your flow?


Please take a try to add a "Filter array" action in your flow replace of the ODATA query, the flow should as below:




Best regards,


Level: Powered On

Re: POSTGRES ODATA time comparison

Using filter contains '2018-09-05' works but since the table has over a million rows the query takes 22 mins to execute if I don’t use the database engine to filter the results.


I will try the filter as you suggested but im sure if it does work the same issue will exist as the table just has too many rows. 

Level: Powered On

Re: POSTGRES ODATA time comparison


Level 8

Re: POSTGRES ODATA time comparison

I'm in the same boat. My research leads me to believe there is no solution. I hope I'm wrong and am still searching...


  • our PG timestamp without time zone columns are considered "DateTime" fields in OData
  • OData intentionally dropped support for DateTime in some prev update (many years ago)
  • OData community is/was livid ( but apparently there's no intent to restore it

I believe our only hope is that Flow will add support for "Execute a query" to on-prem gateways. Here's a link to the community topic. Add your vote!

Level: Powered On

Re: POSTGRES ODATA time comparison

@degvalentineI was able to solve this however it was not an ideal query. You will need to use the MONTH, DAY, and YEAR Functions in combintation with the greate than and less than operators. 


For example if you wanted to compare the date was equal to 04/04/2013

month(DATE) eq 04 AND day(DATE) eq 04 AND year(DATE) eq 2013


Level: Powered On

Re: POSTGRES ODATA time comparison



There are a few aditional options

1 - Create an Azure Function app that would take a HTTP Post with a sql query string and return JSON Object with the query results. Then use an HTTP Connector in flow to pass a SQL Query to the azure function app. 

2 - Use an Azure Logic app instead of flow and you can call an azure function app, or execute some javascript right from the azure logic app. 


Level 8

Re: POSTGRES ODATA time comparison

@ialastairhunter  2 great suggestions - Thank you!


The OData workaround was a bit painful as you suggested, but it works. Doing a "my_ts >= yyyy-mm-dd" filter ended up looking like this (for any that might find this thread later):

year(my_ts) gt yyyy or (year(my_ts) eq yyyy and (month(my_ts) gt mm or (month(my_ts) eq mm and day(my_ts) ge dd)))

I'll try the Azure Function next and post results (and probably code).

Level 8

Re: POSTGRES ODATA time comparison

Couldn't use Azure Functions to access my database because it's on-prem and I don't want to pay for a VPN Gateway.

Helpful resources


Power Automate Community User Group Member Badge

Fill out a quick form to claim your user group badge now!


Incoming: New and improved badges!

We've given our badges an overhaul and also added some brand new ones!


Microsoft Learn

Learn how to build the business apps that you need.


Power Platform World Tour

Find out where you can attend!


Webinars & Video Gallery

Watch & learn from the Power Automate Community Video Gallery!

Top Kudoed Authors
Users Online
Currently online: 443 members 5,901 guests
Please welcome our newest community members: