Data Action - Analytics Export

Voice & Telephony
1

Hey all,

We’ve got a data action pulling historical analytics data - specifically, Contact Summary data - via the /api/v2/analytics/summaries/aggregates/query endpoint. It was working fine, but now it’s choking on rate limits. It’s hitting 429s after about 150 requests per minute. Seems low.

The data action is configured to run every 5 minutes and export data for the last 24 hours. No filtering applied. The script itself is just a simple Python script using requests. We’re not even hitting the documented 600 RPM limit, and we’ve baked in a 2 second delay between calls, so it should be well under. Saw a similar issue last month with a different endpoint - stale OAuth token. This doesn’t feel like that though.

Here’s a sample request body:

import requests
import time

url = "https://api.mypurecloud.com/v2/analytics/summaries/aggregates/query"
headers = {
 "Authorization": "Bearer YOUR_OAUTH_TOKEN",
 "Content-Type": "application/json"
}
data = {
 "interval": "day",
 "metrics": ["totalContacts"],
 "groupBy": ["contactCenter"],
 "timeZone": "America/Chicago"
}

response = requests.post(url, headers=headers, json=data)
print(response.status_code)
print(response.text)
time.sleep(2)

Console logs are mostly the 429s. The error message is standard: “Too Many Requests”. Anyone else seeing this? We’re on genesyscloudrestapi v1.4.1.

2

That rate limit seems kinda low, yeah? We’re on Genesys Cloud too, and I haven’t run into that exact number before.

It’s probably not the script itself, but how often you’re hitting the endpoint. In CIC we used to have the same issue with ICWS - it’s almost always the frequency. That analytics aggregates query is pretty chatty.

Try bumping up the interval on the data action - maybe to 10 or even 15 minutes? Or, even better, break up the 24-hour window. Instead of grabbing everything at once, could you split it into smaller chunks? Like, run the data action four times a day, each grabbing 6 hours of data?

Also - have you checked the historical data action runs? See if you can spot a pattern in the 429s. Sometimes the timing of other processes can make things worse. I usually grab a screenshot of the runs and compare them side-by-side, helps me spot a trend.

If it’s still acting up, you might need to look at the API usage reports. See if there are other processes hammering the analytics API at the same time. It’s a long shot but worth a look.

3

The synchronous endpoint is likely the cause of the rate limit. I would recommend moving to the asynchronous job for these exports.

curl -X POST "https://api.mypurecloud.com/api/v2/analytics/summaries/aggregates/jobs" \
-H "Authorization: Bearer ${TOKEN}" \
-d '{
 "interval": "2023-10-01T00:00:00",
 "metrics": ["n.contact.count"]
}'

The response might be a bit vague, but it’s more stable.

2 Likes
4

That’s correct. Using /api/v2/analytics/summaries/aggregates/jobs is better for large data. It helps when we need high recall for topic trends without hitting the 429 limit.

Sorry if this is a newbie question, but maybe check if the job status is polling too fast. We’ve had some weird errors when checking the jobId too often.

5

the earlier reply is correct. but be careful with job status. you must poll GET /api/v2/analytics/summaries/aggregates/jobs/{jobId} until state is “Completed”. if you call too fast, you get 429 again. just add small sleep in your loop.