Understanding network failures in data centers: measurement, analysis, and implications
By Phillipa Gill, Navendu Jain, N. Nagappan. Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication, 2011
937 citations recorded by the source corpus.
We present the first large-scale analysis of failures in a data center network. Through our analysis, we seek to answer several fundamental questions: which devices/links are most unreliable, what causes failures, how do failures impact network traffic and how effective is network redundancy? We answer these questions using multiple data sources commonly collected by network operators. The key findings of our study are that (1) data center networks show high reliability, (2) commodity switches such as ToRs and AggS are highly reliable, (3) load balancers dominate in terms of failure occurrences with many short-lived software related faults,(4) failures have potential to cause loss of many small packets such as keep alive messages and ACKs, and (5) network redundancy is only 40% effective in reducing the median impact of failure.
Subfields: Computer Networks and Communications, Information Systems
View Paper: https://www.semanticscholar.org/paper/8b43e74d4e3a7584e49073bebd4485049a17f92b
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Networks and Communications
- Mohammad Alizadeh
- Rui Miao
- Vyas Sekar
- George Varghese
- Brighten Godfrey
- Yashar Ganjali
- Amin Firoozshahian
- Cheng Yi
In Information Systems
Citations in the corpus
- SIMPLE-fying middlebox policy enforcement using SDN (2013, background): “Studies show that middlebox overload is a common cause of failures [20, 39], and thus an important consideration is to balance the load across middleboxes.”
- Maglev: A Fast and Reliable Software Network Load Balancer (2016, background): “It has sustained the rapid global growth of Google services, and it also provides network load balancing for Google Cloud Platform.”
- CONGA: distributed congestion-aware load balancing for datacenters (2015, background): “Handling asymmetry essentially requires non-local knowledge about downstream congestion at the switches.”
- DRILL: Micro Load Balancing for Low-latency Data Center Networks (2017, methodology): “Even in large scale datacenters, big groups of correlated link failures are rare with only 10% of failure groups (failures with downtimes either simultaneous or close together in time) containing more than four failures [39].”