Enabled the trace in the UI and ran the API calls. But could not see any trace for the API.
Checked the MP logs, did not see anything there either.
Ran the dig command on the host alias (for ex: www.myhost.com)) used in the URL and got the ELB information.
Found that there were about 6 routers associated with the ELB for this host alias
Checked the ports on which each of the routers were listening.
Found that 2 of the routers were listening on port XXXX which matched with the port # specified in the virtual host
However, the remaining 4 routers were listening on a different port YYYY
Checked the list of routers associated with the specific org/environment combination and found that only the 2 routers (refer to point #6) belonged to this org/environment
This explained why the problem was intermittent. Whenever the API calls hit these 2 routers, we got successful response. When the API calls hit the other 4 routers, we got 404 NOT Found.
Removed the 4 routers which did not belong to this org/environment from the ELB.
With this change, we started seeing successful responses for all the API calls.
@AMAR DEVEGOWDA , Is above issue related to Private Cloud / Public Cloud ? Above issue affected free orgs / paid orgs ? I have seen some community members similar issues. Any details is appreciated.