Digital Garden

Bloom filter is a fast, memory efficient probabilistic data structure that can quickly test whether an element is a member of a set. It can return false positive results, but never false negative results - meaning that if it returns true, the element is pr...

Log-Structured Merge Tree is a Data Structure optimized for write-heavy workloads, used by storage engines like Cassandra and HBase. When a write comes in, we add it to an in-memory balanced tree data structure (for example, a red-black tree), often called...

Let's imagine we are making a simple Key Value Database that writes entries to a file (e.g. Append Only FIle), and we want to implement an in-memory hash map as an index. Storing data on disk Since whenever we update the file we are just appending data...

Declarative query languages provide ways of querying data from Databases by describing the data you want to get. On the other hand, in Imperative Query Languages you query data by describing the way you want to fetch the data. Pros of using declarative que...

Database indexes are data structures that databases keep on the side to improve the speed of reading the primary data, at the cost of increased storage usage and speed of writes. Indexes are derived from the primary data, and in many databases you can safe...

The network Database Model is like an improved Hierarchical Database Model, that allows many-to-many relationships by allowing each node to have multiple parents. Status: #📥 References: Book - Designing Data Intensive Applications (Source)

MySQL is a Relational Database Model Relational Database. Most relational database systems execute the ALTER TABLE statement in a few milliseconds. MySQL is a notable exception—it copies the entire table on ALTER TABLE, which can mean minute...

The hierarchical Database Model organizes data in a tree-like structure, where each record (parent) has one or more child records, but each child has only one parent (in contrast to Network Database Model which allows for multiple parents of a single node)....

Many-to-many relationships are what gives advantage to Relational Database Model over Document Database Model, but in cases when we have a lot of complex many-to-many relationships, it is good to start thinking about how to model our data as a graph. Graph...

Database models are different approaches to structuring and managing data in Databases. It’s not possible to say in general which data model leads to simpler application code; it depends on the kinds of relationships that exist between data items: If th...

Relational Database Model is an approach to structuring and managing data in Databases. Data is stored in tables (also called relations) with predefined schemas (columns with specific data types). Foreign keys define relationships between the tables. ...

Document Database Model is an approach to structuring and managing data in Databases. Data is stored as documents (usually JSON or BSON) inside collections. Documents can be self-contained (denormalized), reducing the need for joins. The main arguments in...

In Relational Database Model Relational Databases, the data is often normalized to reduce redundancy and maintain consistency. Let's say you have a table with student data. If you are storing all their details in the database directly as...

Abstractions are a way of hiding complexity, for example, High-level programming languages are abstractions that hide machine code, CPU registers, and syscalls. When programming in a high-level language, we are still using machine code; we are just not usin...

Anything that can go wrong with software (or hardware) is called a fault. A systems that anticipates faults and can cope with them is called a fault-tolerant or resilient system. It's important to note that faults are always expected and it's impos...

Software failure happens when a system as a whole stops providing expected service to its user. For Reliability, it's important to distinguish between failures and faults (Software Fault). Failures happen as a consequence of faults that haven't bee...

The procedure below can be used in case Storage Class used supports volume expansion, which you can check with: k get storageclass $STORAGE_CLASS_NAME -o yaml | grep allowVolumeExpansion If above command returns true, you can proceed with editing each PVC...

A scalable system is a system that can deal with increasing growth (of data, traffic, complexity) in a reasonable ways. We can never just say that a system is or isn't scalable. The right way to think about scalability is in terms of questions like &qu...

IP IP is a protocol for communication using packets to transmit data. Each IP packet consists of header and data. Packets sent are not guaranteed to arrive in the same order, or to arrive at all - so IP is a best-effort protocol. TCP TCP is a secure pro...

K8S Secrets can't be accessed across different environments. If you want to use an existing secret in another environment, you'll have to copy it there. To copy a secret from one namespace to another, you can use this simple script: SECRET_NAME=$1 ...

The third of The Four Laws of Behavior Change is to make executing Habits easy. This law focuses on making it as easy as possible to start, and keep doing things that pay off in the long run. When you start playing a new instrument, even with a huge amount...

Agaricomycotina is a subphylum of Basidiomycota which holds over 30000 species over 4 classes: Agaricomycetes Bartheletiomycetes Dacrymycetes Tremellomycetes Status: #🌱 References: Video - Professor Dave Explains - Mycology Wikipedia

Agaricomycetes is a fungi class belonging to Basidiomycota Phylum - Taxonomy, Agaricomycotina subphylum. It contains the following Order - Taxonomy: Agaricales Status: #🌱 References:

Genus is the rank in Taxonomy of Living Organisms, that comes after Family - Taxonomy and before Species - Taxonomy. Status: #🌱 References:

In Taxonomy of Living Organisms, any Order - Taxonomy of Fungi has the name suffix -ales. Some of the fungi orders: Agaricales Status: #🌱 References:

In Taxonomy of Living Organisms, any Family - Taxonomy of Fungi has the name suffix -aceae. To see examples of fungi families, see specific Fungi orders. Status: #🌱 References:

Amanitaceae is the name of Fungi family belonging to Agaricales Fungi order. It holds the following Genus - Taxonomy: Amanita Status: #🌱 References:

Agaricales is the name of Fungi order belonging to Basidiomycota Fungi phylum that are characterized by having Gills (Lamellae). It contains 46 families that aren't extinct. Some families in this order are: Amanitaceae Status: #🌱 References: ...

In Biology, Taxonomy is the study of naming and classifying living organisms based on shared characteristics. Organisms are grouped into ranks called taxa (singular taxon), each with increasing specificity: Domain - Taxonomy Kingdom - Taxonomy Phylu...

Phylum is the rank in Taxonomy of Living Organisms, that comes after Kingdom - Taxonomy and before Class - Taxonomy. Each kingdom contains a number of phyla: Animalia: ~31 Plantae: ~14 Fungi: ~8 (see Fungi phylum) Status: #🌱 References: Vide...

Domain is the second-highest rank in Taxonomy of Living Organisms, that comes below Domain - Taxonomy. Domains are divided into smaller groups called Phyla (singular Phylum - Taxonomy). The number of kingdoms of life that exist is different from textbook ...

Kingdom of Fungi has about 8 phyla (Phylum - Taxonomy). They can be divided into 3 groups based on their similarities. All phyla in the fungal kingdom have a name suffix -mycota while subphyla have a suffix -mycotina. Fungi that are zoosporic, single-celle...

Eukarya is one of three Domain - Taxonomy of life in the Taxonomy of Living Organisms. The members of this domain are called eukaryotes. They stand out from the other domains of life by the fact that their cells have a nucleus. All multicellular organisms...

Domain is the highest rank in Taxonomy of Living Organisms. Based on the three-domain-system introduced in 1990, there are three domains of life: Bacteria - Domain Archaea - Domain Eukarya - Domain Status: #🌱 References: Video - Professor Da...

Basidiomycota are one of the Fungi phylum (Phylum - Taxonomy) that belong to Dikarya subkingdom, as they have 2 nuclei per cell. They are considered "higher fungi" along with Ascomycota. The feature that tells them apart from Ascomycota is their ...

Ascomycota are one of the Fungi phylum that belong to Dikarya subkingdom, as they have 2 nuclei per cell. They are considered "higher fungi" along with Basidiomycota. The feature that tells them apart from Basidiomycota is their Hymenium, which i...

Hypha (plural Hyphae) are thin, long filaments forming the Mycelium of Fungi. Status: #🌱 References: Video - Professor Dave Explains - Mycology (source)

Mycology is a branch of Biology studying Fungi. Status: #🌱 References:

Mycelium is a root-like structure of a Fungi which is made out of small filaments called hyphae. It grows through soil or some other substrates. It allows the fungus to grow, absorb nutrients, and produce Mushrooms. Mycelium network can be too small to see...

Mushrooms represent the fruiting bodies of certain types of Fungi that belong to Basidiomycota phylum. They produce spores, which helps the fungus spread. Below the mushroom, inside the soil or some other substrate, a fungus also has a big network of filam...

From top to bottom: Mushroom: Cap (Pileus) Gills (Lamellae) Hymenophore Hymenium Mushroom Spore ...

Fungus (plural Fungi) is any eukaryotic (Eukarya - Domain) organism that belongs to a group of: Yeasts Molds Mushrooms They are specified as a separate Kingdom - Taxonomy of life, distinct from plants and animals. The main characteristic that sets ...

Amanita is a Genus - Taxonomy of Fungi belonging to Amanitaceae family that is characterized by having: Gills (Lamellae) Partial veil (Annulus) Universal Veil (Volva) Some Species - Taxonomy in this genus are: Amanita muscaria Status: #🌱 Re...

Temptation building is a strategy which can be used to increase the attractiveness of a habit by bunding it together with an already-attractive activity. For example, if you love watching a TV show, you could only allow yourself to watch it while (or direc...

When humans started forming tribes to increase their chances of survival, being expelled from a tribe meant certain death. While it's not the case anymore, social influence is still a powerful factor in human life, and we are hardwired to look for ways ...

The second of The Four Laws of Behavior Change is to make Habits more attractive. This law focuses on increasing the attractiveness of desired habits to increase the likelihood of them being completed. We all already have cues that force us toward certain ...

This note serves as a collection of Git-related how-to guides that lead you through the steps of solving real-world problems. To use it, follow the linked mentions below.

Dopamine plays a huge role in forming our desires, and (in this context) can be described as a hormone that forces us to seek pleasure. In addition to this, it also plays a critical role in many important aspects of our bodies like movement, memory, cogniti...

When you amend a Git commit, you will have to force push it to remote, which can cause you to overwrite someone's commit which was pushed in the meantime. Using push flag --force-with-lease instead of --force will not allow you to force push your amend...

Implementation Intention is a strategy that can be used in forming Habits, and talks about having a concrete plan on how you will implement a certain habit into your life. Our habits always start with a cue, and the idea behind implementation intentions is...

Habit Stacking is a strategy which is an expansion of the idea covered in Implementation Intentions, that relies on the fact that habits can be cues to other habits. Once we perform a habit, we can perform another one just after it, and use the momentum we...

Design of our environment is a key component in influencing things we do. Let's see how we can utilize this to form good Habits and break bad ones, using one positive and one negative habit as an example: Let's use taking supplements as a positive ...

In Elasticsearch, modification of cluster state and master node election can only be performed by master nodes. In order to allow for some fault tolerance of the cluster, each decision is made through a quorum of eligible master nodes by getting a confirmat...

StatefulSet is an object that manages a set of K8S Pods and provides guarantees on their ordering and uniqueness. It is pretty similar to K8S Deployment, but has a few important differences which make it a much better candidate for managing stateful applic...

Elasticsearch is a distributed search and analytics engine that serves as a data/document store in Elastic Stack. It stores data in serialized JSON documents which are indexed inside an inverted index data structure to provide very quick lookups. Indexing ...

There are 3 types of autoscaling available in Kubernetes: K8S Cluster Autoscaler Horizontal Pod Autoscaler Vertical Pod Autoscaler Status: #🌱

Kubernetes Cluster Autoscaler represents one of the types of Kubernetes Autoscaling. Its job is to add or remove K8S Node based on cluster usage. It works by checking for K8S Pods that are Pending K8S Scheduling, and creates new Nodes if this is due to ins...

The first of the The Four Laws of Behavior Change is to make cues to your desired habits more obvious. Every Habit starts with a cue (see Habit Feedback Loop), so in order to ensure we have control over our habits, it's essential that we raise awarenes...

A 01 Inbox/Kubernetes Service will expose an IP address, and when you reach it, it will forward your request to one of the underlying pods. This is good if you want to connect to one pod, but what if you want to connect to all of them? This is where the co...

Continuous Improvement is a mindset of applying small, consistent changes in order to achieve big results. For how similar ideas apply to business world, also see Kaizen. It can be a little unintuitive to see why this approach is so effective, so let's...

Four laws of behavior change are principles for designing better Habits described by Atomic Habits Each law of behavior change is tied to one stage of Habit Feedback Loop, telling us how to optimize this part of the habit cycle to reduce the friction and i...

A habit is a behavior that is performed regularly, sometimes automatically, often unconciously. Our brain evolved to "implement" habits to reduce the cognitive load of problems that arise over and over in your environment. You can think of them a...

Habit feedback loop is a model of the mental process that allows habits to be formed and maintained. The habit feedback loop consists of four stages: Cue Craving Response Reward Four stages Reward is the end goal of every habit - it's essen...

Atomic Habits is a system for achieving goals by changing our everyday Habits in small ways, leading to Continuous Improvement. It's based on a Book - Atomic Habits which describes the practical implementation of the system, and talks about what habits...

Rolling Update is a Deployment Strategy (also known as Ramped or Incremental), where we slowly replace version A with version B. How it works before we start, we have 2 instances of A running and serving traffic (let's call this starting point stage...

Recreate Deployment Strategy is a very simple strategy to set up. Here is how it works: you first terminate version A once that is done, you roll out version B The advantage of this approach is that it's really simple to implement. The disadvanta...

Deployment strategies describe how we handle the release of the new code to an environment. Each has it's different use case, pros and cons. Here is a (non-definitive) list of deployment strategies: Recreate Deployment Strategy Rolling Update Depl...

Canary is a Deployment Strategy where we gradually shift traffic from version A to version B. This is usually done via weighted load balancing (e.g., 90% traffic to A, 10% traffic to B). How it works before we start, we have 2 instances of A running and...

Blue/Green is a Deployment Strategy where version B is deployed alongside version A, after which the traffic is switched from A to B. How it works before we start, we have 2 instances of A running and serving traffic we deploy 2 instances of B alongsi...

When designing distributed systems and talking about their consistency and availability, Yield and Harvest are two terms/metrics we can utilize to get a better understanding of how our system behaves, and which tradeoffs we are making Yield Yield is prett...

Network Partition is a division of a network into a separate subnets, either by design or by network failure. When a network partition occurs, all traffic sent from a component in partition A to components in partition B will be lost. Any pattern of messag...

Databases are computer systems designed for storing data. This note serves as a link to connect database-related notes. ACID Optimistic Locking/Pessimistic Locking Database Index Status: #🌱 Tags: #🗺️

CAP theorem states that distributed data stores can provide only two of the Consistency or Availability and Partition Tolerance. Since network failures are unavoidable, we can assume that Partition Tolerance is something we need to have. CAP theorem states ...

Unix Philosophy is a set of concepts that pushes the creation of software that is modular and reusable. It is summarized by a couple of key ideas: Write programs that do one thing and do it well. Write programs to work together. Write programs to ha...

Single Responsibility Principle (OCP) is the first of SOLID Principles which states: "A class should have one, and only one, reason to change" In other words, if some code has to change, it should be for its own reasons, not as collateral damage...

Pessimistic Locking in Database systems is a locking mechanism that relies on exclusive locks to ensure two transactions don't update rows at the same time. For example: user A fetches account data which says amount: 100 and acquires shared read loc...

Optimistic Locking in Database systems is a locking mechanism that relies on data versions to stop updates based on stale data. For example: user A fetches account data which says amount: 100, version: 1 user B fetches the same row, getting amount: 10...

ACID is a set of properties of Database systems that talks about how a database transaction should behave in order for it to be considered reliable. It stands for: Atomicity Consistency Isolation Durability Atomicity Atomicity says that either ...

SOLID principles are five design principles created to improve object-oriented Software Design. Even though its primary focus is OOP, a lot can be learned from these principles in terms of non-OOP languages like 01 Inbox/Golang. The five principles are: ...

Open/Closed Principle (OCP) is the second of SOLID Principles which states: "Software entities should be open for extension, but closed for modification" In a bit clearer wording, you should be able to extend the behavior of a system, without ha...

In Linux, whenever a user application wants to touch any of these: files any hardware or devices running processes networking stuff date/time writing to stdout it can't do so on its own - it must ask Linux Kernel to do it instead. Syscall...

Tracing is a type of Monitoring which sacrifices the number of events it looks at to give us a picture of how a system behaves. It's heavily relying on sampling - e.g. looking at only every hundredth request. Because of this, it can provide more contex...

Prometheus is an open source, metrics based Monitoring system. Its data model is kept as a time series, each consisting of key value pairs called labels. PromQL is a querying language that allows for aggregation across any of these labels, allowing you to ...

Prometheus stores data locally on the disk in a custom Database. It doesn't support any form of clustering by itself, in an attempt to make running Prometheus a simple task. This means that you are limited by the storage space on your machine. This als...

Once your code is instrumented, Prometheus needs a way to find the services which are exposing the metrics. Yes, you could tell Prometheus where the services are and where they expose metrics, but this approach is not too scalable. Prometheus has integrati...

Recording rules allow us to evaluate and store results of PromQL queries in storage as a new time series, so we don't have to calculate everything on the fly whenever we want to render a graph. They are run on regular intervals to update the resulting t...

Prometheus has a number of http apis which allow you to request raw data and evaluate PromQL queries. Often needed or computationally expensive queries can be optimized by the use of Prometheus Recording Rules. Status: #🌱 References: Book - Promethe...

Once Prometheus Service Discovery gets us the list of targets to be monitored, Prometheus fetches the metrics by sending a "scrape" http request. Once Prometheus receives a response to the scrape request, it parses the response, enriches it with ...

Prometheus can create alert based on Alerting Rules. Alerting rules are another form of Prometheus Recording Rules, which turn the results of the PromQL expressions into alerts that are sent to Alertmanager. Alertmanager receives alerts from Prometheus and...

Profiling is a type of Monitoring where we collect a lot of data for a short period of time, which we can use to debug an issue. Profiling allows us to have all the events with most of the context for a limited period of time. It can generate a ridiculous ...

Logging is a form of Monitoring which looks at a limited set of events (e.g. all http requests, all db queries) and records part of their contexts Logging sacrifices the amount of context provided for each event, but avoids sampling, so it's a perfect ...

Depending on which software you are trying to expose Prometheus metrics for, there are three options to achieve this. Pre-instrumented software Many applications are already exposing metrics in Prometheus format. In case you are monitoring an application ...

Different layers of infrastructure and application are exposing the same Monitoring metrics. For example, your Database reports the query duration, and so does your application. These two are sure to have similar results – on application side, add a couple ...

In statistics, n-th percentile represents the value below which n% records can be found. It's calculated by sorting a data set and dividing it into 100 equal groups, and then dividing the same data set into two parts where the lower part contains n% of ...

Traffic shows us how much demand is placed on the system, and is one of the The Four Golden Signals of Monitoring. How you track traffic, largely depends on what type of service you are running. For web services, traffic can be represented as a number of ...

Saturation is a representation of how full your service is, and is one of The Four Golden Signals of Monitoring. When thinking about saturation, it's useful to put the emphasis on the most constrained resource for your application – If memory is our bi...

Response time is the time taken to serve a request, and is one of The Four Golden Signals of Monitoring. It's often synonymously used with the term Request Latency, but it's good to note that response time is what the client sees, while latency is j...

Errors show us a rate of failed requests, and are one of The Four Golden Signals of Monitoring. Error rates can look at requests that failed: directly (5XX errors) – fetched at load balancer level is an easy way to track this globally. See Where to Coll...

White Box Monitoring is when we monitor the internal workings of our system. For example, users have no idea about our current CPU utilization, so that metric is a White Box metric. Its primary use is in collecting telemetry and debugging, but also has an i...

When setting up Alerting for the first time, many people instinctively set alerts on Cause Based Monitoring metrics – if my service's CPU is ramped to 100%, of course I want to be alerted! If someone asks "why is that so?", a simple answer is...

Symptom Based Monitoring points us to allows us to observe the user experience. A metric is Symptom based if it shows an actual symptom that is making our users happy or sad. We gather Symptom based metrics almost exclusively through Black Box Monitoring. S...

Cause Based Monitoring points us to a cause of an existing issue, but don't imply that issue exists in the first place. Some examples of Cause Metrics are: CPU utilization Free disk space When users are seeing slow response times, I want to be ab...

Black Box Monitoring is when we look at our system from the perspective of our users – without knowing anything about its internal state. Since Black Box Monitoring is looking at customer experience, it can show us any currently active problems (e.g., high...

When setting up Monitoring for a system that is able to output A LOT of metrics, it's often hard to determine which of these we should be monitoring. If not sure where to start, covering The Four Golden Signals of Monitoring is typically a good place t...

The four golden signals of Monitoring are: Measuring Request Latency Measuring Traffic Measuring Error Rate Measuring Service Saturation Status: #🌲 References: Book - Site Reliability Engineering (Source)

Monitoring is an integral part of running services in production. Without it, we are blind to what's going on, and thus unable to act according to our best interest. Providing visibility is in the center of monitoring, but speaking more broadly, monito...

This note serves as a link to connect Game Development related notes. Status: #🗺️

In Linux Networking, the most basic form of a network is two PCs connected with a switch. In order for machines to connect to a switch, we need an interface on each host (these can be physical or virtual). To see the interfaces on the host, we can run the ...

A router helps connect two networks together. It can be thought of as another host with many network ports. Since it connects to two networks, it will be assigned two IPs - one on each network. In order for a system to reach a system in another network, it...

This note serves as a link to connect Linux Networking related notes. Get started on the topic by checking one of these out: How to make a Linux Host act as a Router Some of the useful networking commands: # list and modify interfaces on the host ip l...

Let's say we have 3 hosts (A, B and C), and two networks (192.168.1.0 and 192.168.2.0): Host A is connected to network 192.168.1.0 via interface eth0 and has IP 192.168.1.5 Host C is connected network 192.168.2.0 via interface eth0 and has IP 192.1...

Insights gained through Meditation practice is one of the main things which leads to Awakening. Insight into the true nature of reality that radically transforms our understanding of ourselves and our relationship to the world is called Vipassana. In addi...

Mundane insights are useful insights gained through Meditation practice which teach us about: our own personality social interactions human behavior how everyday world works Note that while mundane insights are very useful, if they don't impa...

Peripheral Awareness makes sure that while our Attention is focused on a single thing, we are still generally aware of the things in the background. Peripheral awareness is more holistic, open and inclusive than Attention. It provides context for our Model...

The model of Conscious Experience is a basic way to represent how our mind works, in order to be able to better understand it and more easily talk about it. Conscious Experience takes two forms of interacting with Consciousness which are crucial to differe...

"The beauty and significance of life well lived consists not in the works we leave behind, or in what history has to say about us. It comes from the quality of concious experience that infuses our every waking moment, and from the impact we have on oth...

Our consciousness consists of whatever we are experiencing in the moment. It's built from the things we percieve with our senses, as well as mental objects like thoughts, feelings and memories. Status: #🌱 References:

Whenever Attention is focused on some object, that object dominates our Consciousness. Attention singles out a small part of the field of Consciousness, and deeply analyzes and interprets it. Together with Peripheral Awareness, Attention works to form the...

CORS (Cross-Origin Resource Sharing) is a mechanism that allows the server to dictate from which other origins a browser should be allowed to load its resources. For example, if i serve a website from my-domain.com, and use javascript to try to fetch a res...

Axis-Aligned Bounding Box is one of the simplest collision detection methods in 2d Game Development, which only works on 2d rectangles that aren't rotated (their sides are parallel/perpendicular). AABB works by checking if sides of rectangle A are outs...

When you pass an argument to a function in 01 Inbox/Golang, the function will behave differently depending on the type of the argument passed. Here we make a diffrerence between Value Types and GO Reference Types With Value Types, go makes a copy of the va...

When you pass an argument to a function in 01 Inbox/Golang, the function will behave differently depending on the type of the argument passed. Here we make a diffrerence between GO Value Types an Reference Types. With Reference Types, go passes the referen...

Work performed by SRE teams can be divided into a few types: Engineering work Overhead Toil Engineering work Engineering work means both software engineering and systems engineering. Engineering work is what makes real progress on the project, so...

Toil is a type of work which is manual, repetitive, and brings no real long-term value to the project. In addition to this, toil is work that scales linearly with service, and can be solved by automation. If some of these conditions are met, you are likely...

To get the value of a K8S Secret, all you need to know is the secret name and the name of the key that you want to read. What I would normally do before is: kubectl get secret mysecret -o yaml copy the (often very long) base64 string under the key i w...

Service Level Objectives are values (or ranges of values) in which SLIs are allowed to be. For example, if SLI is request latency, SLO could be that request latency should be less than 100ms. SLOs are important in many ways in SRE philosophy, so it's i...

The first thing to know when choosing SLIs and SLOs is that SLOs should always be defined first. The thing we want to avoid by this is just picking whatever's easy to measure and ending up with useless objectives. When defining SLOs, we should always s...

When deciding the level of availability we want for our services, the target that we want to achieve is often described as a percentage of time the service is available. It's worth noting that 100% availability, while it sounds like a nice thing to hav...

Site Reliability Engineering is what happens when you treat operations as a software engineering problem. The goal of SRE is to make software reliable and scalable. SRE can be seen as a specific implementation of DevOps. Some notes on this topic to check o...

Service Level Indicators are quantitative measures of provided level of service, often aggregated into rates, averages, percentiles. Common SLIs are availability, error rate, latency, throughput, durability of data. We have a lot of metrics we are Monitor...

A reliable system is a system that continues to work correctly at the desired level of performance even when faced with issues like hardware and Software Faults or human errors. A unreliable system on the other hand, does not tolerate faults, and lets them...

In order to know how available your service is, you need a way to measure it. One of the most straightforward ways to measure this is by measuring uptime: availability = uptime / (uptime + downtime) In some cases (e.g. global systems), a better availabil...

It's difficult for product and ops teams to find middle ground between investing in reliability vs taking risks. If you test your software too much before releasing, you are going too slow and the market will swallow you, but if you don't test enoug...

Stake Social darwinists were (mostly) a group of sociologists who tried to apply Darwinism to social issues, and mostly to answer the question of how the idea of evolution could explain the social world around them. One of the most prominent social darwin...

Natural selection is a process which allows the evolution of species to better suit the life in their environment, and is the basis of Darwinism. It's based on the fact that organsims that are better suited for life in the environment will survive more ...

Differential survival happens when different organsims of the same species must compete for what they want (reproduction, food, shelter…). It's an important factor in Natural Selection because it makes only the best fitted organsims pass down their gene...

Darwinism is a theory of biological evolution developed by Charles Darwin which describes that species are adapting to their environment through a process of Natural Selection. Status: #🌱

Parmenides was a Pre-Socratic Philosophers who went against the grain and whose ideas were completely oposite to everyone else's at the time. While philosophers have started to view universe as an ever-changing system (Everything is in permanent state ...

Node Selectors are used to place K8S Pods on specific K8S Node by label. For more complex placement options, see K8S Node Affinity. It's possible to make a pod go to a specific node based on the node label. For example, if you have one large node and t...

Node Affinity is a feature which ensures that K8S Pod end up running on specific K8S Node. We are already able to do this with K8S Node Selectors, but with big limitations. See also K8S Anti-affinity if you need to set affinity based on running pods and th...

While K8S Node Affinity works like an extended K8S Node Selectors, Inter-pod affinity and Anti-affinity focus on labels of the pods rather than labels of the K8S Node. This allows us to dictate which pods can or can't be deployed together on the same n...

Git is a version control system used to track changes on files and make it easier for multiple people to collaborate on the same set of files. This note serves as a link to connect Git-related notes. Status: #🗺️

To completely delete a commit from a Git branch you can use the following command: CURRENT_OS=$(uname -s) if [ "$CURRENT_OS" = "Linux" ]; then git rebase --rebase-merges --onto $1^ $1 fi if [ "$CURRENT_OS" = "Darwin...

I switch machines i work on often, and its important to me that my configuration is up-to-date on all of them. I needed a solution which would address the following topics: dotfiles should be backed up in one place which will be the source of truth i ...

A bare Git repository is a repository without a working tree. If you create a bare repository, you'll get just the contents that you'd normally find within .git directory. Since it doesn't have a working tree, the bare repository does not hold ...

In case you are running a Kubernetes cluster in a HA setup with say, 3 K8S Master Nodes, loss of majority of the nodes will stop the cluster from reaching Consensus and thus, from functoning. The main reason this breaks tha HA cluster is that etcd cluster ...

Static pods are K8S Pods run solely by Kubelet on any K8S Node without any interference from kube-apiserver and other controlplane components. Kubelet uses /etc/kubernetes/manifests directory to host static pod manifests by default (directory is configurab...

Consensus is an important concept in Distributed Systems, which allows multiple servers to agree on some information. In order for consensus algorithms to work, there are multiple rules which need to be followed by its members: if a process is to write ...

This page acts as a map of content for Software Design topics. Status: #🗺️

Composition Over Inheritance as OOP principle doesn't have much value if we don't invest time to think about the cases when Inheritance should be favored over Composition. If a class would expose all public methods of another class, it's a good...

Liskov Substitution Principle is the third of the Software Design SOLID Principles in OOP. It states this: If S is a subtype of T, then objects of type T may be replaced with objects of type S without breaking the program.1 Or in plain english - if we hav...

Composition is a method of writing OOP code. It attempts to solve the same Software Design problem as Inheritance, but instead of relying on inheriting features from the parent class, composition focuses on embedding objects inside of objects to allow the s...

Pushdown Automata is an implementation of a Finite State Machine Design Pattern which uses a Stack Data Structure to store it's states. The top of the stack represents the currently active state, while the one below it is the state that we just transit...

Linear interpolation (commonly refered to as lerp) is a useful function in the fields of Game Development and Creative Coding. It's used to get a number on a specific point between two numbers. The parameters are the start and end of the desired range,...

One of the most well-known Design Patterns in Software Design. It's useful when we see a system having a number of different states it can be in, with specific rules of which states can be accessed from which states. It's called a finite state machi...

In order to enable encryption of data stored in etcd cluster, K8S Apiserver needs to be configured with the --encryption-provider-config flag: First, create a file encryption-config.yaml: apiVersion: apiserver.config.k8s.io/v1 kind: EncryptionConfiguratio...

Not many facts were known about Pythagoras, and most stuff we know today is full of speculation. He was sort of a self-proclaimed math-loving cult-running messiah who loved math and Astronomy. He set up a commune in italy where around 300 philosophers join...

Pre-Socratics are philosophers who developed original schools of thought and predated @Socrates. It's important to note that division to before and after Socrates comes not as much from the importance of Socrates' ideas, as from the fact that Philo...

Philosophy is a term coined by Pythagoras meaning "love of wisdom". This note serves as a Map of Content for the Philosophy topic. Follow these notes to explore this topc: Pre-Socratic Philosophers Status: #🌱 Tags: #🗺️

Godfather of the idea that everything is made of atoms and empty space. The guy who turned Monism into a taboo doctrine. It is said that he came up the idea of atoms by searching for a solution to the Arrow Paradox. He recognized that the paradox is unsolv...

Etcd is a reliable key-value Database. It is one of the most important parts of a Kubernetes cluster used to store the data of the K8S Apiserver. It uses Raft Protocol to establish Consensus between its nodes. Status: #🌱 References:

OG philosopher, genious businessman, and an overall smart guy. There's a story where he earned a lot of money by predicting how olive market will behave, which he had figured out because of the climate that year. He basically lived in a world where eve...

Secrets are K8S Objects which are used to store sensitive information you can't otherwise put into a Pod spec or a K8S ConfigMap. Secrets are very similar to ConfigMaps, the biggest difference being that they are designed to be used for confidential da...

Kubernetes scheduler is a K8S Master Node component in charge of K8S Scheduling. Status: #🌱 References:

Grumpy guy going around the street insulting everyone. Due to his behavior, he didn't have many followers to tell about gim, and his notes aren't that revealing as well as he was purposfully cryptic to "hide his thoughts". His teachings t...

Kubernetes Worker nodes are the place where the actual work is being executed inside k8s cluster. They run workloads as dictated by the K8S Master Node, and update the master node on their status so it can manage them. These are the main components of the ...

Service Accounts are K8S Objects used to help pods with K8S Authentication. The ServiceAccount object is namespace-scoped. It's used to allow a process in a pod to access Kubernetes API Server. There exists a default service account object, but it has ...

Replication Controller is part of K8S Controller Manager in Kubernetes. It is responsible for monitoring the status of K8S Replicasets, and ensures that the desired number of pods are available at all times. If a K8S Pod dies, it creates another one. Sta...

ReplicaSets are K8S Objects used to make sure a certain number of K8S Pods is running on the Kubernetes cluster. ReplicaSet uses the labels defined in the selector property to find all the pods with the specified labels. Once it finds them, it makes sure t...

Pod is a K8S Object used to run one or more tightly coupled containers and is rarely used directly, especially in production because of the limitations on the number of properties that can be updated. Instead, pods are usually managed by other objects like ...

To see a list of available Kubernetes objects on your cluster, you can run: # all kubectl api-resources # only namespaced kubectl api-resources --namespaced=true # only cluster-scoped kubectl api-resources --namespaced=false Links to some of the obje...

Node Controller is responsible for monitoring the status of the K8S Nodes and keeping them in the desired state. It does this through the K8S Apiserver. This is the workflow by which Node Controller works: If a new node is created, Node Controller will ...

Kubernetes Master nodes (also known as k8s controlplane) are K8S Nodes responsible for managing the cluster. No user processes normally run on master node, this is the job of the K8S Worker Nodes. These are the main components of the k8s controlplane: e...

Kube Controller Manager is a K8S Master Node component that manages various controllers in Kubernetes. Controllers are processes which continiously monitor the status of various parts of the cluster, and remediate the situation if that part of the cluster i...

Docker provides us a way to map ports of a container to ports of the host (docker run -p 9000:80 nginx) . This will make sure that any traffic that reaches host on the port 8080, will be forwarded to our container on port 80. Docker accomplishes this with ...

CIDR (Classless Inter-Domain Routing) notation can be used to represent a range of IP addresses. It has two parts: prefix - Network address (a normal IP address) suffix - number of bits of the network address provided in the prefix An example cidr bl...

Once you run kubectl run busybox --image busybox on your Kubernetes cluster, what has to happen before your K8S Pod is running successfully? The K8S Apiserver is the component which recieves your request, authenticates you, validates the request and writes...

Kubernetes is a container orchestration platform. This note serves as a Map of Content for this topic. Start your exploration on one of these notes: K8S Node K8S Object What happens when you create a Pod in Kubernetes Status: #🌲 Tags: #🗺️

Kubelet runs on a Kubernetes node and is responsible for managing the node it's runnning on. It starts and stops nodes as requested by the K8S Apiserver. It also updates the kube apiserver on the worker node status in regular intervals. Kubelet registe...

K8S Scheduler is making sure that every K8S Pod is assigned to a K8S Node. Every pod that gets created has a property NodeName which is not set by default. kube-scheduler looks for pods with these fields unset and assigns them a node by setting the NodeName...

Kubernetes implements CNI(Container Network Interface) to allow third-party networking solutions to integrate with it. Each CNI solution needs to implement a set of things including, but not limited to: creating bridge networks attaching containers to...

Kube Apiserver is the primary component of a Kubernetes cluster. It runs on a K8S Master Node, and is the only component which is talking directly to the etcd cluster. It exposes a REST API through which all K8S components interact with the cluster / with e...

Docker is a Container Engine that helps people easily develop and ship their applications. This note serves as a link to connect Docker-related notes. Docker Networking Status: #🌲 Tags: #🗺️

When running a Docker container, there are several networking options to chose from: None Runs a container in a completely isolated network. Container can't reach anything outside of it, and nothing can reach the container. docker run --network none n...

Whenever a container is created, Docker creates a 01 Inbox/Network Namespaces for it. You wont be able to see it when running ip netns because Docker doesn't create a symlink for it. Let's assume you created the symlink as described in the link. i...

When Docker is installed, it creates an internal private network which is by default called bridge . You can see this network like this: docker network ls #> NETWORK ID NAME DRIVER SCOPE #> 71f0f1f94055 bridge bridge local...

CoreDNS is one of many DNS Server solutions, which is particuarily interesting because it's used for DNS Resolution in Kubernetes. Most of its functionality is implemented through plugins, so it's very flexible and extensible by nature. Let's s...

DNS Server is a server who's primary purpose is the DNS resolution. It holds a Database of IP addresses and Domain Names, and is able to communicate with other DNS Servers to help with the queries it doesn't know about. They use port 53 (thus Route5...

DNS Resolution is a process in which Domain Name are converted to IP addresses of the hosts they point to. The process of resolution is handled by multiple DNS servers which forward the requests to each other in order to find the server with the actual IP ...

To figure out how DNS Resolution works in Kubernetes, there are a few important components to consider: k8s DNS server running on the cluster (e.g. CoreDNS) K8S Node's local DNS resolution (e.g. /etc/resolv.conf) Kubelet which creates pods and pr...

This is a quick walkthrough of how I manage and publish my notes. Goal The goal of this whole project is for me to have a good way to manage my knowledge base and easily publish some subset of it. The notes should be stored in markdown format in Git, and...

While our local network hosts can be referenced with names such as db, on the internet, hosts use a special Domain Name format: www.google.com. The purpose of this format is to group similar things together. The last portion of the domain (.com) is called ...

DNS (Domain Name System) provides us a way to give names to ip addresses. For example, if you have a Database running on a host in your local network behind the ip 192.168.1.6 (Host B), you can name it "db" by adding a line in the hosts file on yo...

There are multiple record types that can be stored in a DNS server: IP -> Hostname is known as A Record IPv6 -> Hostname is known as AAAA Record Hostname -> Hostname is known as C Record Status: #🌱