Automatic recovery from runtime failures

Abstract

We present a technique to make applications resilient to failures. This technique is intended to maintain a faulty application functional in the field while the developers work on permanent and radical fixes. We target field failures in applications built on reusable components. In particular, the technique exploits the intrinsic redundancy of those components by identifying workarounds consisting of alternative uses of the faulty components that avoid the failure. The technique is currently implemented for Java applications but makes little or no assumptions about the nature of the application, and works without interrupting the execution flow of the application and without restarting its components. We demonstrate and evaluate this technique on four mid-size applications and two popular libraries of reusable components affected by real and seeded faults. In these cases the technique is effective, maintaining the application fully functional with between 19% and 48% of the failure-causing faults, depending on the application. The experiments also show that the technique incurs an acceptable runtime overhead in all cases.

Are mutants a valid substitute for real faults in software testing?

“…Mutants are valuable for this reason and commonly used (e.g., [4,19]). Yet, it is unclear whether those algorithms evaluated on mutants perform equally well on real faults.…”

Automatic error elimination by horizontal code transfer across multiple applications

“…Rx [50] and ARMOR [20] are runtime recovery systems based on periodic checkpoints. When an error occurs, Rx [50] reverts back to a previous checkpoint and makes system-level changes (e.g, thread scheduling, memory allocations, etc.)…”

“…to search for executions that do not trigger the error. ARMOR [20] reverts back to a previous checkpoint and finds semantically equivalent workarounds for the failed component based on user-provided specifications.…”

Beyond the rainbow: self-adaptive failure avoidance in configurable systems

“…Weimer et al [44] finds faults in systems using test cases, and then uses an extended form of genetic programming to fix the code. Carzaniga et al [2][3][4], find workarounds which are code-level alternate library calls. This technique is similar to ours in its goal and the requirements to find a workaround.…”

“…Most work on self-adaptation has focused on continuous quality attributes allowing techniques from feedbackcontrol to be applied [6,9,12,16,20,24,40]. Recent work by Carzaniga et al [2] uses self-adaptation in a different way -to avoid executing faulty code after a failure is observed. Their code adaptation automates a kind of workaroundavoiding rather than repairing the fault by modifying the code to use alternative libraries.…”