DiMOS (greek δημος = people) is a Distributed Metrics Observation System. It contains facilities to acquire metrics in a reliable way, aggregate them, filter outliers, vote on diverging metrics, and visualise the filtering and voting processes. DiMOS can be used across different application areas, as collaborative research infrastructure or as prototypical solution in commercial use cases.
Application Areas
AI
In data-centric artificial intelligence, emphasis is put on the quality of input data due to its impact on autonomous decision-making processes. DiMOS is a good fit for preprocessing input data in distributed environments.
IoT
High-quality sensing requires redundant measurement points from which the relevant values need to be determined. DiMOS offers a suitable foundation for bringing more data science into the Internet of Things.
Cloud Computing
Cloud software artefacts are known to require heavy testing to avoid inconsistencies, quality deficiencies and bugs. DiMOS offers methods and tools to automate such testing, assessment and insights generation.
Data Science
Data scientists can use DiMOS to reduce data multiplicity and ensure data consistency according to specified rules. DiMOS is also integrated with Renku to facilitate reproducible data science workflows involving consensus voting.
Software Tools
DiMOS Orchestrator (formerly MAO Orchestrator)
The orchestrator schedules metrics acquisition and observation through tools executed in sequential workflows (pipelines). It performs generic actions to increase reliability, including rescheduling in cases of failure, outlier detection, and voting.

»» Git
Visualisation
A web-based visualisation of multiple independent nodes in a federation of metrics observers including timeline and statistics on agreements and disagreements.
Acquisition and Insights Tools
Tools are containerised software responsible for either acquisition or aggregation of metrics, or both. Their functionality depends on the data sources (e.g. websites, APIs or sensors).

»» JSON API Tools
»» JSON API Pipelines
VDX: Voting Definition Schema
VDX: An all-purpose system specification for data-centric sotware voters. It aims to be a simple way to define the behavior of voting software, and helps to purify AI input data from multiple sources.

»» Git
Observations
Microservice Artefact Observatory
The 'original' community of researchers performing static and dynamic analysis of microservice artefacts such as Helm charts, AWS SAR, Dockerfiles and Docker images among others.

»» Website
MuSDaFA
Researchers in 'Multi-Sensor Data Fusion with Accretion' use DiMOS to vote over multiple sensor data streams or beacon signals.

»» Website
More observations: Registry of federations
DiMOS is increasingly useful for different domains. The orchestrator's installer makes it easy to join an existing federation. They can also be inspected beforehand.

»» JSON API Federations
Publications
Decentralised Data Quality Control in Ground Truth Production for Autonomic Decisions
IEEE TPDS, accepted for publication, 2022
Autonomic decision-making based on rules and metrics is inevitably on the rise in distributed software systems. Often, the metrics are acquired from system observations such as static checks and runtime traces. To avoid bias propagation and hence reduce wrong decisions in increasingly autonomous systems due to poor observation data quality, multiple independent observers can exchange their findings and produce a majority-accepted, complete and outlier-cleaned ground truth in the form of consensus-supported metrics. In this work, we motivate the growing importance of metrics for informed and autonomic decisions in clouds and other distributed systems, present reasons for diverging observations, and describe a federated approach to produce ground truth with data-centric consensus voting for more reliable decision making processes. We validate the system design with experiments in the area of cloud software artefact observations and highlight benefits for reproducible distributed system behaviour.

»» Zenodo
»» Git
Towards Reproducible Software Studies with MAO and Renku
SoftwareX, accepted for publication, 2022 In software engineering, the developers' joy of decomposing and recomposing microservice-based applications has led to an enormous wave of microservice artefact technologies. To understand them better, researchers perform hundreds of experiments and empirical studies on them each year. Improving the reuse and reproducibility of these studies requires two ingredients: A system to automate repetitive experiments, and a research data management system with emphasis on making research reproducible. Both frameworks are now available via the Microservice Artefact Observatory (MAO) and Renku. In this paper, we explain the current capabilities of MAO as a global federated research infrastructure for determining software quality characteristics. Moreover, we emphasise the integration of MAO with Renku to demonstrate how a reproducible end-to-end experiment workflow involving globally distributed research teams looks like.

...
DiMOS is mainly researched and developed at Zurich University of Applied Sciences, Switzerland, along with research partners. This website is work in progress. Revisit later to get more information -- or ask!