Performance evaluation of a distributed storage service in community network clouds

Abstract

Community networks are decentralized communication networks built and operated by citizens, for citizens. The consolidation of todays cloud technologies offers now for community networks the possibility to collectively built community clouds, building upon user-provided networks and extending towards cloud services. Cloud storage and in particular secure and reliable cloud storage, could become a key community cloud service to enable end-user applications. In this paper we evaluate in a real deployment the performance of Tahoe-LAFS, a decentralized storage system with provider-independent security that guarantees privacy to the users. We evaluate how the Tahoe-LAFS storage system performs when it is deployed over distributed community cloud nodes in a real community network. Furthermore, we evaluate Tahoe-LAFS in the Microsoft Azure commercial cloud platform, in order to compare and understand the impact of homogeneous network and hardware resources on the performance of the Tahoe-LAFS. We observed that the write operation of Tahoe-LAFS resulted in similar performance when using either the community network cloud or the commercial cloud, but the read operation achieved better performance in the Azure cloud, where the reading from multiple nodes of Tahoe-LAFS benefited from the homogeneity of the network and nodes. Our results suggest that Tahoe-LAFS can run on community network clouds with suitable performance for the needed end-user experience.

Cloudy in guifi.net: Establishing and sustaining a community cloud as open commons

“…To ensure a satisfactory user experience, each application was carefully tested and some were tweaked (e.g., PeerStreamer) before becoming part of the pre-installed applications of the Cloudy software stack [55]. Assessments were done for Tahoe-LAFS [56], PeerStreamer [57], and the search service [58]. The evaluations included the deployment of the applications under realistic conditions on several nodes in the CN, considering the relevant metrics affecting the user experience in each case (e.g., read times in the case of Tahoe-LAFS, video chunk losses in the case of PeerStreamer, and service discovery effectiveness in the case of the search service).…”

Gelly-Scheduling: Distributed Graph Processing for Network Service Placement in Community Networks

“…We believe guifi.net, while it is in fact an open community network, has a type of topology which allows for extrapolating results into other sorts of networks. This claim is made based on previous research work in the literature [14], which both analyzed the impact of prioritizing different heuristics on the computational and network resources available [16] and studied practical issues with micro-service architectures [13]. We used available statistical processing tools to attempt to fit several distributions and compare them.…”

“…Nodes in guifi.net are exclusive to specific geographical zones (there are no overlays) such as what is depicted in Figure 1. There are special-purpose nodes called graph-servers, which are responsible for performing network measurements between nodes and have an API for querying node states [14]. These graph-servers comprise a distributed hierarchical monitoring system which records the network's link data traffic properties.…”

A Lightweight Service Placement Approach for Community Network Micro-Clouds

“…A micro-cloud is nothing but a platform to deliver services to a local community of citizens within the vast CN. Services can be of any type, ranging from personal storage [29] to video streaming and P2P-TV [28]. Observe that this model is different from Fog computing [9,21], which extends cloud computing by introducing an intermediate layer between devices and datacenters.…”