Dan Alistarh

Fast and Exact Majority in Population Protocols

Population protocols, roughly defined as systems consisting of large numbers of simple identical agents, interacting at random and updating their state following simple rules, are an important research topic at the intersection of distributed computing and biology. One of the fundamental tasks that a population protocol may solve is majority: each node starts in one of two states; the goal is for all nodes to reach a correct consensus on which of the two states was initially the majority. Despite considerable research effort, known protocols for this problem are either exact but slow (taking linear parallel time to converge), or fast but approximate (with non-zero probability of error). In this paper, we show that this trade-off between precision and speed is not inherent. We present a new protocol called Average and Conquer (AVC) that solves majority exactly in expected parallel convergence time O(log n/(sε) + log n log s), where n is the number of nodes, ε n is the initial node advantage of the majority state, and s = Ω(log n log log n) is the number of states the protocol employs. This shows that the majority problem can be solved exactly in time poly-logarithmic in n, provided that the memory per node is s = Ω(1/ε + log n log{1/ε). On the negative side, we establish a lower bound of Ω(1/ε) on the expected parallel convergence time for the case of four memory states per node, and a lower bound of Ω(log n) parallel time for protocols using any number of memory states per node.

StackTrack: an automated transactional approach to concurrent memory reclamation

Dynamic memory reclamation is arguably the biggest open problem in concurrent data structure design: all known solutions induce high overhead, or must be customized to the specific data structure by the programmer, or both. This paper presents StackTrack, the first concurrent memory reclamation scheme that can be applied automatically by a compiler, while maintaining efficiency. StackTrack eliminates most of the expensive bookkeeping required for memory reclamation by leveraging the power of hardware transactional memory (HTM) in a new way: it tracks thread variables dynamically, and in an atomic fashion. This effectively makes all memory references visible without having threads pay the overhead of writing out this information. Our empirical results show that this new approach matches or outperforms prior, non-automated, techniques.

Fast randomized test-and-set and renaming

Most people believe that renaming is easy: simply choose a name at random; if more than one process selects the same name, then try again. We highlight the issues that occur when trying to implement such a scheme and shed new light on the read-write complexity of randomized renaming in an asynchronous environment. At the heart of our new perspective stands an adaptive implementation of a randomized test-and-set object, that has poly-logarithmic step complexity per operation, with high probability. Interestingly, our implementation is anonymous, as it does not require process identifiers. Based on this implementation, we present two new randomized renaming algorithms. The first ensures a tight namespace of n names using O(n log4 n) total steps, with high probability. This significantly improves on the complexity of the best previously known namespace-optimal algorithms. The second algorithm achieves a namespace of size k(1 + e) using O(k log4 k/ log2(1 + e)) total steps, both with high probability, where k is the total contention in the execution. It is the first adaptive randomized renaming algorithm, and it improves on existing deterministic solutions by providing a smaller namespace, and by lowering step complexity.

Sub-logarithmic test-and-set against aweak adversary

A randomized implementation is given of a test-and-set register with O(log log n) individual step complexity and O(n) total step complexity against an oblivious adversary. The implementation is linearizable and multi-shot, and shows an exponential complexity improvement over previous solutions designed to work against a strong adversary.

The Complexity of Renaming

We study the complexity of renaming, a fundamental problem in distributed computing in which a set of processes need to pick distinct names from a given namespace. We prove an individual lower bound of \Omega( k ) process steps for deterministic renaming into any namespace of size sub-exponential in k, where k is the number of participants. This bound is tight: it draws an exponential separation between deterministic and randomized solutions, and implies new tight bounds for deterministic fetch-and-increment registers, queues and stacks. The proof of the bound is interesting in its own right, for it relies on the first reduction from renaming to another fundamental problem in distributed computing: mutual exclusion. We complement our individual bound with a global lower bound of \Omega( k \log ( k / c ) ) on the total step complexity of renaming into a namespace of size ck, for any c \geq 1. This applies to randomized algorithms against a strong adversary, and helps derive new global lower bounds for randomized approximate counter and fetch-and-increment implementations, all tight within logarithmic factors.

Optimal-time adaptive strong renaming, with applications to counting

We give two new randomized algorithms for strong renaming, both of which work against an adaptive adversary in asynchronous shared memory. The first uses repeated sampling over a sequence of arrays of decreasing size to assign unique names to each of n processes with step complexity O(log3 n). The second transforms any sorting network into a strong adaptive renaming protocol, with an expected cost equal to the depth of the sorting network. Using an AKS sorting network, this gives a strong adaptive renaming algorithm with step complexity O(log k), where k is the contention in the current execution. We show this to be optimal based on a classic lower bound of Jayanti. We also show that any such strong renaming protocol can be used to build a monotone-consistent counter with logarithmic step complexity (at the cost of adding a max register) or a linearizable fetch-and-increment register (at the cost of increasing the step complexity by a logarithmic factor).

Are lock-free concurrent algorithms practically wait-free?

Lock-free concurrent algorithms guarantee that some concurrent operation will always make progress in a finite number of steps. Yet programmers prefer to treat concurrent code as if it were wait-free, guaranteeing that all operations always make progress. Unfortunately, designing wait-free algorithms is generally a very complex task, and the resulting algorithms are not always efficient. While obtaining efficient wait-free algorithms has been a long-time goal for the theory community, most non-blocking commercial code is only lock-free. This paper suggests a simple solution to this problem. We show that, for a large class of lock-free algorithms, under scheduling conditions which approximate those found in commercial hardware architectures, lock-free algorithms behave as if they are wait-free. In other words, programmers can keep on designing simple lock-free algorithms instead of complex wait-free ones, and in practice, they will get wait-free progress. Our main contribution is a new way of analyzing a general class of lock-free algorithms under a stochastic scheduler. Our analysis relates the individual performance of processes with the global performance of the system using Markov chain lifting between a complex per-process chain and a simpler system progress chain. We show that lock-free algorithms are not only wait-free with probability 1, but that in fact a general subset of lock-free algorithms can be closely bounded in terms of the average number of steps required until an operation completes. To the best of our knowledge, this is the first attempt to analyze progress conditions, typically stated in relation to a worst case adversary, in a stochastic model capturing their expected asymptotic behavior.

Tight Bounds for Asynchronous Renaming

This article presents the first tight bounds on the time complexity of shared-memory renaming, a fundamental problem in distributed computing in which a set of processes need to pick distinct identifiers from a small namespace. We first prove an individual lower bound of Ω(k) process steps for deterministic renaming into any namespace of size subexponential in k, where k is the number of participants. The bound is tight: it draws an exponential separation between deterministic and randomized solutions, and implies new tight bounds for deterministic concurrent fetch-and-increment counters, queues, and stacks. The proof is based on a new reduction from renaming to another fundamental problem in distributed computing: mutual exclusion. We complement this individual bound with a global lower bound of Ω(k log (k/c)) on the total step complexity of renaming into a namespace of size ck, for any c ≥ 1. This result applies to randomized algorithms against a strong adversary, and helps derive new global lower bounds for randomized approximate counter implementations, that are tight within logarithmic factors. On the algorithmic side, we give a protocol that transforms any sorting network into a randomized strong adaptive renaming algorithm, with expected cost equal to the depth of the sorting network. This gives a tight adaptive renaming algorithm with expected step complexity O(log k), where k is the contention in the current execution. This algorithm is the first to achieve sublinear time, and it is time-optimal as per our randomized lower bound. Finally, we use this renaming protocol to build monotone-consistent counters with logarithmic step complexity and linearizable fetch-and-increment registers with polylogarithmic cost.

How to Solve Consensus in the Smallest Window of Synchrony

This paper addresses the following question: what is the minimum-sized synchronous window needed to solve consensus in an otherwise asynchronous system? In answer to this question, we present the first optimally-resilient algorithm ASAPthat solves consensus as soon as possiblein an eventually synchronous system, i.e., a system that from some time GSTonwards, delivers messages in a timely fashion. ASAPguarantees that, in an execution with at most ffailures, every process decides no later than round GST+ f+ 2, which is optimal.

How efficient can gossip be? (on the cost of resilient information exchange)

Gossip, also known as epidemic dissemination, is becoming an increasingly popular technique in distributed systems. Yet, it has remained a partially open question: how robust are such protocols? We consider a natural extension of the random phone-call model (introduced by Karp et al. [1]), and we analyze two different notions of robustness: the ability to tolerate adaptive failures, and the ability to tolerate oblivious failures. For adaptive failures, we present a new gossip protocol, TrickleGossip, which achieves near-optimal O(n log3 n) message complexity. To the best of our knowledge, this is the first epidemic-style protocol that can tolerate adaptive failures. We also show a direct relation between resilience and message complexity, demonstrating that gossip protocols which tolerate a large number of adaptive failures need to use a super-linear number of messages with high probability. For oblivious failures, we present a new gossip protocol, CoordinatedGossip, that achieves optimal O(n) message complexity. This protocol makes novel use of the universe reduction technique to limit the message complexity.