Sales (Private)

Internet Performance Alerts with Traceroutes

Learn all you need to know about email best practices, deliverability, and tools with email whitepapers and ebooks.

Issue link: https://hub.dyn.com/i/1017032

Contents of this Issue

Navigation

Page 1 of 2

Before - for root cause identification : {"city": "Washington", "reply_ttl": 245, "mpls_labels": [],"adminsubdiv": "District of Columbia", "ip": "4.53.116.42", "label": "Level 3 Communications, Inc.", "rtt": 70.136, "country_code": "US", "asn": 3356, "admindiv": "DC" }, {"city": "Dulles Town Center", "reply_ttl": 59, "mpls_labels": [],"adminsubdiv": "Loudoun County", "ip": "69.59.224.84", "label": "Vonage Holdings,Inc.", "rtt": 70.453, "country_code": "US", "asn": 17151, "admindiv": "VA" } After - for root cause identification . Notice that the link that was previously less than 1ms is now over 20ms in the alert phase. "city": "Washington", "reply_ttl": 245, "mpls_labels": [],"adminsubdiv": "District of Columbia", "ip": "4.53.116.42", "label": "Level 3 Communications, Inc.", "rtt": 69.86, "country_code": "US", "asn": 3356, "admindiv": "DC", "city": "Dulles Town Center", "reply_ttl": 59, "mpls_labels": []," adminsubdiv": "Loudoun County", "ip": "69.59.224.84", "label": "Vonage Holdings, Inc.", "rtt": 93.383, "country_code": "US", "asn": 17151, "admindiv": "VA" Frequently Asked Questions What are you monitoring? Monitoring is configured to run against any IP address configured by the user. How are you monitoring the target IP addresses? We use PING to monitor the target IP addresses, which uses ICMP echo requests and receive back ICMP echo reply messages. How frequently are the PINGs sent? PINGs are sent every 30-45 seconds from a given Vantage Point. Is the monitoring going to add unnecessary load on my infrastructure? No. The frequency of our measurements is extremely low relative to internet traffic. We are deliberately using a known networking standard protocol (ICMP) so that it can be managed by typical networking traffic shaping, if necessary. Most networks apply filters to prevent a massive flood of ICMP traffic, by rate limiting ICMP. With more capacity and resilient infrastructure, this isn't a major network concern any more. And, given our small amount of traffic, we do not see this issue affecting our monitoring or analysis. ICMP provides us with the consistency we need to assess performance over time, as well as the ability to identify when performance deteriorates or fails completely. Traffic engineering will drop our packets if a machine is under load, and we will identify failures indicating that load. Our measurements provide a safe way of working with existing network engineering techniques to achieve reliable monitoring and performance insight. What are the criteria for raising alerts? There are three types of performance alerts: • Connectivity • Packet loss • Latency deterioration The user can configure thresholds for latency and @dyn dyn.com

Articles in this issue

Links on this page

view archives of Sales (Private) - Internet Performance Alerts with Traceroutes