A data-center monitoring platform should pull interface counters from the switches through their API instead of scraping each switch web page. Which approach is that?
Select an answer to reveal the explanation.
Short Explanation
You’re right to replace web-page scraping with API integration, so your monitoring platform can pull interface counters directly through the switches’ published interface. SNMP polling can collect the counters but is not the requested API call; continuing to scrape keeps the method you’re replacing, and syslog provides log lines rather than a pull of those counters.
Full Explanation
API integration is the approach that lets your platform request interface counters through the switches’ published interface and receive structured data instead of scraping a web page. Because collection no longer depends on the page’s appearance, a cosmetic layout change will not disrupt it. SNMP polling is a tempting alternative because it can collect the same counters, but it uses a different management protocol rather than the API the platform was instructed to call. Continuing to scrape each switch’s web page retains the very method the platform is supposed to stop using. A syslog feed supplies log lines; it is not a pull of interface counters through the API. Exam caveat: use a read-only API account for this job. If the platform is compromised, an account that can change switch settings creates a greater risk than one limited to reading counters. Practical check: compare the API’s counters for one interface with the counters shown on the switch for that same interface, and confirm that changing the web-page layout no longer affects collection. Also make a test API call and verify that it cannot change a switch setting before retiring the scraper.