Posts

I was planning to turn my RedisHack/Kaggle competition project into set of smaller open source projects for several months now. Here is the first one: RedisGears based NLP pipeline https://github.com/applied-knowledge-systems/the-pattern-platform — turn text into knowledge graph (stored in RedisGraph) using Medical Ontology (methathesaurus).

Good article Sharath, in the algorithm presented above one step is missing - weights need to be normalised during exploitation step after for loop.. This dramatically improve the stability of the particle filter.

Greate article! I wish I found it earlier: I am on final legs of configuring gentoo/glusterfs (over sd cards, fast usb sticks and slow usb drives) , consul, nomad. I would like to use your PiHole setup, but how do you configure your router if pihole can be dynamically deployed using nomad job?

Over the past few weekends, I have been distracted: I looked at Kubernetes storage in the past (got minio installed on Libreelec), and recently Longthorn announced GA — the project I am keen to explore.

Good article, just as hint for other generations you can check 50-cloud-init.yaml by $ sudo netplan –debug try $ sudo netplan –debug generate $ sudo netplan –debug apply see Q&A

This was Redis #beyongcache #hackathon submission: The whole pipeline is open-sourced in https://github.com/AlexMikhalev/cord19redisknowledgegraph Vision Search or rather information exploration should be spatial preferably in VR (memory palace, see Theatre of Giulio Camillo).

Initially, I only planned to use Redis Cluster as one large in-memory store/database, store all article ids in the set corresponding to the pipeline step and then run: for each article_id in a set, apply next step in the pipeline, save id into next set.

Since learning that I could not use RedisBloom (also RedisSearch and RedisGraph) on the open-source cluster, I replaced the use of Bloom/Cockoo filter with set — (SADD, SMEMBERS), supported since Redis version 1.

In the last article, I span up Redis Cluster. This time is to populate it with data. This time I am not going to create a one large data science pipeline, which takes 9 GB RAM just to start, but I am going to split it into “nano” services, which I can debug easily (and hopefully add tests :) at some point.

First, I already had a two stopped Azure instances and ansible inventory file: cat myazure_rm.yaml plugin: azure_rm include_vm_resource_groups: - cord19demo auth_source: cli keyed_groups: - prefix: tag key: tags(base) Two other instances I created manually by hands and corresponding inventory file: