DEP-LS 2015 - 1st workshop on dependability in large-scale systems

The probability of failures grows proportionally with the size of the system. Large-scale systems exhibit a short mean time between failures (MTBF); the MTBF of exa-scale systems is expected not to exceed one hour. Furthermore many large-scale systems, like clouds for instance, service many users concurrently and thus face highly dynamic workloads. Many kinds of failures can occur in such systems. Physical hosts or virtual machines (VMs) can crash, network partitions or VM migrations can lead to message losses or large variations in network latency, so on and so forth.
Dependability in large-scale systems raises many challenging issues, such as failure detection among a great number of nodes and the design of dependable systems able to deal with high dynamism and virtualization. Combining scalability and dependability also complexifies experimentation: it is hard to control a vast experimental environment where deployment, fault injection, and performance analysis are conducted on a large scale.
This workshop aims to bring together researchers from different domains to create a forum that investigates the many aspects of dependability in large-scale and dynamic systems. Contributions may span a broad range of topics, including but not limited to:
Failure detection
Churn resilience
Checkpointing
Reliable data storage
Information dissemination
Robustness evaluation
Failure injection
DEP-LS 2015 is held in conjunction with the eleventh European Dependable Computing Conference.