Sending AWS CloudWatch Alarms to ServiceNow
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
02-28-2023 02:32 AM
Does anyone have any idea how to use AWS Cloudwatch push connector in servicenow event management
How to send information from AWS Cloudwatch to servicenow event table
I created a topic in AWS
And used the following endpoint - https://myinstancename.service-now.com/api/global/em/inbound_event?source=aws.
After creating a subscription, it asks me to confirm subscription.
How I can confirm subscription,
It was asking me subscription confirmation url. Where I can find this Url.
And in which json format I need to send information.
whether I can use the same json format they are sending or I need to ask them to change the json format to map with event table .
Can it directly map with event table if they don't change it ?

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-01-2023 09:37 PM
Refer below URLs for detailed Steps.
Regards
RP

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
06-15-2023 12:59 AM - edited 06-15-2023 01:02 AM
The endpoint needs to include username and password. In the documentation the API is different from Utah than Tokyo. I have only made it work with the API mentioned in the Tokyo version, even in my own Utah dev instance:
https://<username>:<password>@<instance-name>.service-now.com/api/global/em/inbound_event?source=aws
Username and password connects to a user in SN that you create with the the evt_mgmt_integration role.
If you set this up correct the subscription confirmation should be done automatically. There is no confirmation URL.
Also: only Cloudwatch Alarms will work, not Cloudwatch Events created through event rules. This is because events doesn't include the header x-amz-sns-message-type with value Notification that is needed for the push connector to handle incoming events. This is easily overlooked and we spent hours troubleshooting before noticing this.