GFCache: A Greedy Failure Cache Considering Failure Recency and Failure Frequency for an Erasure-Coded Storage System

In the big data era,data unavailability,either temporary or permanent,becomes a normal occurrence on a daily basis.Unlike the permanent data failure,which is fixed through a background job,temporarily unavailable data is recovered on-the-fly to serve the ongoing read request.However,those newly revived data is discarded after serving the request,due to the assumption that data experiencing temporary failures could come back alive later.Such disposal of failure data prevents the sharing of failure information among clients,and leads to many unnecessary data recovery processes,(e.g.caused by either recurring unavailability of a data or multiple data failures in one stripe),thereby straining system performance.To this end,this paper proposes GFCache to cache corrupted data for the dual purposes of failure information sharing and eliminating unnecessary data recovery processes.GFCache employs a greedy caching approach of opportunism to promote not only the failed data,but also sequential failure-likely data in the same stripe.Additionally,GFCache includes a FARC(Failure ARC)catch replacement algorithm,which features a balanced consideration of failure recency,frequency to accommodate data corruption with good hit ratio.The stored data in GFCache is able to support fast read of the normal data access.Furthermore,since GFCache is a generic failure cache,it can be used anywhere erasure coding is deployed with any specific coding schemes and parameters.Evaluations show that GFCache achieves good hit ratio with our sophisticated caching algorithm and manages to significantly boost system performance by reducing unnecessary data recoveries with vulnerable data in the cache.

An efficient and highly available framework of data recency enhancement for eventually consistent data stores

Data items are usually replicated in modem dis- tributed data stores to obtain high performance and avail- ability. However, the availability-consistency and latency- consistency trade-offs exist in data replication, thus system designers intend to choose weak consistency models, such as eventual consistency, which may result in stale reads. Since stale data items may lead to serious application semantic problems, we consider how to increase the probability of data recency which provides a uniform view on recent versions of data items for all clients. In this work, we propose HARP, a framework that can enhance data recency of eventually con- sistent distributed data stores in an efficient and highly avail- able way. Through detecting possible stale reads under fail- ures or not, HARP can perform reread operations to elim- inate stale results only when needed based on our analysis on write/read processes. We also present solutions on how to deal with some practical anomalies in HARP, including de- layed, reordered and dropped messages and clock drift, and show how to extend HARP to multiple datacenters. Finally we implement HARP based on Cassandra, and the experi- ments show that HARP can effectively eliminate stale reads, with a low overhead （less than 6.9%） compared with original eventually consistent Cassandra.

Improvising Personalized Travel Recommendation System with Recency Effects

A travel recommendation system based on social media activity provides a customized place of interest to accommodate user-specific needs and preferences. In general, the user’s inclination towards travel destinations is subject to change over time. In this project, we have analyzed users’ twitter data, as well as their friends and followers in a timely fashion to understand recent travel interest. A machine learning classifier identifies tweets relevant to travel. The travel tweets are then used to obtain personalized travel recommendations. Unlike most of the personalized recommendation systems, our proposed model takes into account a user’s most recent interest by incorporating time-sensitive recency weight into the model. Our proposed model has outperformed the existing personalized place of interest recommendation model, and the overall accuracy is 75.23%.

Evaluation of the HIV Case-Based Surveillance System: A Pilot of the Electronic Health Record System in Mutare District, Zimbabwe, 2021

Background: Zimbabwe started HIV case-based surveillance in April 2017. Rapid testing for HIV recent infection was introduced into routine HIV and testing services in 2019 along with the Impilo Electronic Health Record System. For the period January-June 2020, only 1 out of 13 health facilities in Mutare district reported seven newly diagnosed HIV patients through the electronic health record system compared to 483 in the District Health Information System (DHIS-2) recorded from paper-based registers. We evaluated the case-based surveillance system attributes, usefulness and reasons for under-reporting from January-December 2020. Methods: We conducted a descriptive cross-sectional study using updated Centres for Disease Control guidelines for evaluating public health surveillance systems. Questionnaires were administered to 36 health workers involved in HIV testing services. Facility checklists were used to collect data on knowledge, system attributes and usefulness of the system. Completed HIV case-based surveillance forms were assessed for completeness. Epi Info Version 7 was used to generate frequencies, means and proportions. Results: The reasons for under-reporting of patients in the electronic health record system were lack of reporting guidelines 26/36 (72%), limited coordination between technical staff and health facilities 24/36 (67%) and limited competency on the Electronic health record system 22/36 (61%). Timeliness, completeness, and validity were 88%, 82% and 100% respectively. The stability of the system was affected by the lack of standard operating procedures during system interruptions. Overall representativeness was 45% despite increasing from 3/226 (1%) to 224/303 (73%) between Quarter-1 and Quarter-4 of 2020. Acceptability was 100% due to reduced paperwork and the ability to generate simple reports. The information generated was used to identify new infection hotspots 28/36 (78%). Conclusion: The HIV cases based surveillance system was timely, acceptable with good data quality. Representativeness was poor due to limited competency on the electronic health record system. As a result, health workers received further training.