k8S ❤️ KServe

2026-05-04T01:01:35+00:00

</a> Serving ML Models on Kubernetes with KServe</h1>
Deploying a trained model is often harder than training it. You need an API endpoint, autoscaling, health checks, request batching, and a way to handle both lightweight predictive models and heavyweight generative ones — without writing bespoke infrastructure for each. KServe solves this by giving you a standard, Kubernetes-native way to serve models of any kind.
This post walks through what KServe is, how it’s architected, and how to install it and deploy your first models.

`kubectl</code> — configured with access to a Kubernetes cluster</li>`
`helm</code> (v3+)</li>`
`A Kubernetes cluster (local via kind/minikube, or a managed cluster)</li> </ul>`</a> What Is Model Serving?</h2> Model serving is the process of taking a trained model and making it usable in production. At a high level, it involves: Loading the model into memory so it’s ready to handle requests</li> Exposing an API endpoint (HTTP or gRPC) that clients can call</li> Handling the request lifecycle — preprocessing input, tokenization, running inference, and postprocessing output</li> Managing operational concerns — batching concurrent requests, health checks, and graceful restarts/rollouts Serving needs differ significantly depending on the type of model:</li> </ol> </th> Predictive Models</th> Generative Models</th></tr></thead> Output</td> A single output per input (e.g. a class label, a score)</td> A stream of tokens, generated one at a time until an end token</td></tr> Resource usage</td> Typically lightweight, CPU-friendly</td> Compute-heavy, usually GPU-bound, and latency-sensitive due to streaming</td></tr> </tbody></table> This distinction matters because a serving platform built only for one type (say, small scikit-learn models) won’t scale well to something like an LLM — and vice versa. </a> Why KServe?</h2> A few realities make model serving on Kubernetes genuinely hard to do by hand: Models are resource-hungry. GPUs and large memory footprints are the norm, especially for generative models.</li> Scale is unpredictable. Traffic can spike or drop suddenly, and you need to scale (including to/from zero) without manual intervention.</li> You rarely run just one model. Most real deployments serve many models, often of different frameworks and formats.</li> Open-weight models don’t simplify serving. Downloading a model from Hugging Face is the easy part — production-grade serving (batching, routing, autoscaling, observability) is still on you. KServe addresses this by providing a Kubernetes Custom Resource Definition (CRD), InferenceService</code>, that abstracts away the deployment details. You describe what you want to serve; KServe handles how.</li> </ul> </a> Kserve Architecture</h2> KServe sits on top of Kubernetes and (optionally) a service mesh/ingress layer, and is built around a controller that reconciles InferenceService</code> resources into running deployments. Core components: </a> Installation</h2> KServe is installed via Helm in three layers: CRDs, the core controller, and runtime configs. # kserve-crds:: Core KServe CRDs ( InferenceService, TrainedModel ) helm install kserve-crd oci://ghcr.io/kserve/charts/kserve-crd --version v0.19.0 # custom resource definitions:: LLM-specific LLMInferenceService helm install kserve-llmisvc-crd oci://ghcr.io/kserve/charts/kserve-llmisvc-crd --version v0.19.0 # install kserve # -- no Istio Service mesh or ingress controller:: avoids requiring Istio/Knative — runs as plain Kubernetes Deployments helm install kserve oci://ghcr.io/kserve/charts/kserve-resources --version v0.19.0 --namespace kserve --create-namespace --set kserve.controller.deploymentMode=RawDeployment</code></pre> </a> Verify the controller is up</h4> # Check Status kubectl rollout status deployment/kserve-controller-manager -n kserve # install runtime configs # - servingruntime.enabled=true ;; Serving runtimes are opt-in helm install kserve-runtime-configs oci://ghcr.io/kserve/charts/kserve-runtime-configs \ --version v0.19.0 \ --namespace kserve \ --set kserve.servingruntime.enabled=true \ --set kserve.controller.gateway.disableIstioVirtualHost=true \ --set kserve.controller.gateway.disableIngressCreation=true helm status kserve-runtime-configs -n kserve helm get all kserve-runtime-configs -n kserve kubectl get pods -n kserve kubectl get crds | grep kserve kubectl get clusterservingruntimes</code></pre> </a> Deploying Your First Models</h2> Namespace</li> </ol> kserve-inference</code> namespace created. # kserve_ns.yaml apiVersion: v1 kind: Namespace metadata: name: kserve-inference</code></pre> Qwen2 model as generative-model example.</li> </ol> # qwen2_small.yaml apiVersion: serving.kserve.io/v1beta1 kind: InferenceService metadata: name: qwen-model namespace: kserve-inference spec: predictor: model: modelFormat: name: huggingface storageUri: "hf://Qwen/Qwen2.5-0.5B-Instruct" args: - --backend=huggingface resources: requests: cpu: "1" memory: "2Gi" limits: cpu: "2" memory: "6Gi"</code></pre> Sklearn Iris model as predictive model example.</li> </ol> # iris.yaml apiVersion: serving.kserve.io/v1beta1 kind: InferenceService metadata: name: sklearn-iris namespace: kserve-inference spec: predictor: model: modelFormat: name: sklearn storageUri: https://storage.googleapis.com/kfserving-examples/models/sklearn/1.0/model/model.joblib</code></pre># bert-classifier.yaml apiVersion: serving.kserve.io/v1beta1 kind: InferenceService metadata: name: sentiment-classifier spec: predictor: model: modelFormat: name: huggingface storageUri: hf://distilbert/distilbert-base-uncased-finetuned-sst-2-english args: - --model_name=sentiment-classifier - --task=sequence_classification - --backend=huggingface - --return_probabilities resources: requests: cpu: 1 memory: 2Gi limits: cpu: 2 memory: 4Gi </code></pre> The KServe controller watches InferenceService</code> objects, picks the matching ClusterServingRuntime</code> for the model format, provisions the necessary compute, wires up networking, and reports status back — following the standard Kubernetes reconciliation pattern. # Wait for the model to become ready kubectl wait --for=condition=Ready --timeout=10m inferenceservice/sklearn-iris -n kserve-test kubectl describe pod/PODSNAME -n kserve-test kubectl get deployment,service -n kserve-test # port forwarding kubectl port-forward -n kserve-test svc/sklearn-iris-predictor 8080:80 # check kubectl get inferenceservice qwen-model -n kserve-inference kubectl get pods -n kserve-inference -w</code></pre> </a> Calling the Inference API</h2> KServe’s generative model runtimes expose an OpenAI-compatible chat completions endpoint: # Generative POST /openai/v1/chat/completion - modelname - list_of_messages - max_tokens # Predictive POST /v2/models/MODEL_NAME/infer - name - shape - data-type - data # call api curl -s http://localhost:8080/openai/v1/chat/completions -H "Content-Type: application/json" -d '{"model":"qwen-model", "messages":"{"role":"user","content":"What is Kserve?"}", "max_tokens":100 }' | jq curl -s http://localhost:8080/v2/models/sentiment-classifier/infer -H "Content-Type: application/json" -d '{"inputs": [{"name": "input-0","shape": [1],"datatype": "BYTES","data": This course was delightful. The labs taught me how to serve ML models and query them."]}]}' | jq .</code></pre> </a> Uninstall</h2> helm uninstall kserve --namespace kserve helm uninstall kserve-runtime-configs --namespace kserve helm uninstall kserve-crd helm uninstall kserve-llmisvc-crd</code></pre> </a> Conclusion</h2> KServe gives you a single, declarative interface (InferenceService</code>) for serving both lightweight predictive models and heavyweight generative ones on Kubernetes — without hand-rolling autoscaling, batching, or health-check logic yourself. Once the controller and runtimes are installed, deploying a new model is often just a matter of writing a short YAML manifest pointing at a model artifact. My Love ❤️ 2025-01-17T00:00:01+00:00 எண்ணம் உண்டோ என்னை👉👈 அறிய கண்ணில் மை பூசி காதல்🫶🏼 செய்ய தாய்மை அறியவே இனி தனிமை தவிர்ப்பேன் இனி நீ தஞ்சம் புகும் புதுமணையில்🏡 வர்ணம்🌈 தீட்டு வானை🦚 அறிந்து வளையல் பூட்டு, காதல்❤️ பேசு காவியம் பாடு▷ , புதுமை படி அதை எளிமை செய் ஆனந்தம்🎼 பாடு அதில் அர்த்தம் அறி என்ன என்று எடுத்துறை காது பிடித்து கடிந்துறை 😤 ஊருக்கு உறைக்கும் முன் என்னிடம் உண்மை சொல் நான் ஓய்ந்து இருந்தால் ஒழுக்கம் செய் அயர்ந்து இருந்தால் அள்ளிக்கொள் 🤗 வரதட்சணையாக வாரம் 12 முத்தங்களும், வழக்காட சில குறும்புகளும் - வந்திரனும் சீர்வரிசை யாக சில சில சண்டைகள் போதும் அறிவேனோ ஆயிரம் நான் அறை பொழுதிலே ஆனால் அமரன்‌‌👑 போல வாழ்வேன் என்றும் என் மேல் அன்பு🫰 கொள்வியானாள் Tamil Tokenizer 2024-01-09T05:20:35+00:00 Project Url</a></a> MediaToAscii 2023-07-01T05:20:35+00:00 Project:: https://github.com/Muthukamalan/media2ascii Git in Practical - Intermediate 2023-06-22T21:01:35+00:00 Table Of Content</summary> </details> </a> Introduction</h2> Distributed version control systems combine the capabilities of local and remote systems. Developers can store the entire codebase locally, allowing independent work and offline changes. Developers then synchronize their local changes to a central server for other collaborators to consume. where as, GitHub is a cloud-based service that hosts Git repositories (Centralized Platform) </a> Installation</h3> # GIT sudo apt update && sudo apt install git git --version # GIT LFS sudo apt-get install git-lfs git lfs install</code></pre> </a> Configuration</h3> config file loaded fron ~/.gitconfig</code> or ~/.config/git/</code> dir [user] name = USER_NAME email = USER_MAIL [core] editor = code --wait [init] defaultbranch = main [log] decorate = full [rerere] enabled = false [advice] forcedeletebranch = true [alias] lol= "!git --no-pager log --oneline --decorate --graph --parents" wdiff = diff -w --word-diff=color --ignore-space-at-eol ss = status -bsv [commit] verbose=true</code></pre> </a> Global/Local Namespace</h3> --global adding config in global space</code> --local adding config in local git-space level</code> git config --global user.name "Your name" git config --global user.email "email@domain.com" git config --global core.editor "code --wait" git config --global init.defaultbranch main git config --global log.decorate full git config --global rerere.enabled false git config --global advice.forcedeletebranch true # set git config --add --local webflyx.tone Black git config --add --local webflyx.status Married git config --add --local webflyx.status FamilyMan # unset git config --unset --local webflyx.tone # unset-all git config --unset-all --local webflyx.status # remove section git config --remove-section webflyx</code></pre> </a> Git Tracks content</h4> Git is a content tracking system</code>. The Git Object store is based on the hashed</code> computation of the content</code> of its object not on the file or dir names Example, If two seperate file located in 2 different dir have exactly the same content, Git stores a sole copy of that content as blob within the object store</li> It stores every version of every file - not the differences.</li> </ul> System</th> Index Mechanism</th> Data Store</th></tr></thead> Data Base</td> Indexed Sequential Access Method</td> Data Records</td></tr> Git</td> .git/objects/hash</code>, tree object contents</code></td> Blob Objects, tree objects</td></tr> </tbody></table> .git/ ├── branches ├── config ├── description ├── HEAD ├── hooks │ ├── applypatch-msg.sample │ ├── commit-msg.sample │ ├── fsmonitor-watchman.sample │ ├── post-update.sample │ ├── pre-applypatch.sample │ ├── pre-commit.sample │ ├── pre-merge-commit.sample │ ├── prepare-commit-msg.sample │ ├── pre-push.sample │ ├── pre-rebase.sample │ ├── pre-receive.sample │ ├── push-to-checkout.sample │ ├── sendemail-validate.sample │ └── update.sample ├── info │ └── exclude ├── objects │ ├── info │__ └── pack └── refs ├── heads └── tags</code></pre> </a> initalize</h3> git init touch README.md && git add . git commit -m 'Initial commit' git clone <url> git clone -b --single-branch rbranch <url> # git branch -a && git checkout <remote-branch> git clone <url> <dir> # Clone particular folder in the Repo git clone --depth=1 <url> # bring only one commit git clone --mirror <repository-url> # mirror # Adding to repo git remote add origin <url> git remote -v # details about push/pull git remote set-url origin <new-url> # origin sets upto new url git push -u <origin/REMOTE-NAME> <local-branch-name>:<remote/origin> # -u make it upstream git push -d <branch> # delete the branch</code></pre> </a> Log</h3> git --no-pager log --oneline --decorate --graph --parents --all git show <commit-hash> # Examine commit git show HEAD</code></pre> </a> branching</h3> git switch <existing-branch> git switch -c <new-branch> git branch git branch -a git branch -r git branch -d <branch-name> # keep git history git branch -D <branch-name> # remove git history</code></pre> </a> update only the latest commit</h3> git commit --amend -m 'Updated commit message' # update only the message # Forget file in the commit git add <file(s)> git commit --amend --no-edit # Undo the Last commit git reset --soft HEAD~1 # --soft option is useful if u want to go back to prev commit, but keep all of your changes. Committed changes will be uncommited and staged, while uncommitted changes will remain staged or unstaged as before. git reset --hard HEAD~1 # delete it from the staging area as well.</code></pre> </a> Tag</h3> git tag # list all tags git show tag-no. git tag -a tag-no # with annotated-message git push --tags # push only tags git push origin tag-no # push only that tag git tag tag-no commit-hash git tag -d tag-noo git tag -a v1.0.0 -m 'Release 1.0.0' # RECOMMENDED WAY git push origin v1.0.0 # push specific tag git push origin --tags # push all tags git push origin --delete v1.0.0 # delete remote tag</code></pre> </a> Merge</h3> conflict markers <<<<<<< HEAD Changes introduced by the main branch ====== Changes introduced by the widgets branch >>>>>>> branch</code></pre>git merge --abort git switch <main> && git merge branch # --ff --no-ff --ff-only git merge --squash/--no-squash # in the branch do squash and merge</code></pre> </a> Fetch</h3> # git push uploads commits from a local repository to a remote repository, while git fetch downloads commits. # git pull = git fetch + merge git fetch origin main -v git switch main git merge origin/main # git rebase origin/main</code></pre> </a> stash</h3> When you are in the middle of something, your boss comes in and demands that you fix something immediately. Traditionally, you would make a commit to a temporary branch to store your changes away, and return to your original branch to make the emergency fix, like this: - ... hack hack hack ... - $ git switch -c my_wip - $ git commit -a -m "WIP" - $ git switch master - $ edit emergency fix - $ git commit -a -m "Fix in a hurry" - $ git switch my_wip - $ git reset --soft HEAD^ - ... continue hacking ...</code></pre>- ... hack hack hack ... - $ git stash - $ edit emergency fix - $ git commit -a -m "Fix in a hurry" - $ git stash pop - ... continue hacking ...</code></pre>git stash -u # stash untracked file as well -m "" git stash show git stash pop git stash apply stash@{2} # git stash pop --index 2 git stash drop stash@{2} git stash clear # clear all</code></pre> </a> Revert</h3> creates a new commit that undoes a previous one — safe for shared/remote branches since it doesn’t rewrite history git revert <commit-hash> # undo a specific commit (opens editor) git revert <commit-hash> --no-edit # skip the editor, use default message git revert HEAD # undo the last commit git revert HEAD~3..HEAD # revert a range (newest to oldest) git revert -n <commit-hash> # stage the revert without committing (--no-commit) git revert --continue # after resolving revert conflicts git revert --abort # abandon the revert mid-way</code></pre> Note: If you pushed changed been revert, no merge conflict will happen</code> </a> Reset</h3> Reset current HEAD to the specified state Note: Never use git reset on commits already pushed</code> git reset --hard COMMISH # clears the mess from the index file and the working tree. git reset --soft COMMISH # leaves all your changed files</code></pre> </a> Restore</h3> git restore is a focused command for discarding working tree changes or unstaging files # Discard working directory changes (unstaged) git restore <file> # restore file to last commit state;; Git will retrieve the state of these file(s) from the HEAD commit, and revert them back to that version git restore . # restore all modified files # Unstage a file (move from index back to working tree) git restore --staged <file> # you accidentally add file to stage then wanna remove;; equivalent to: git reset HEAD <file> git restore --staged . # unstage everything # Restore from a specific source git restore --source <commit-hash> <file> # restore file from any commit git restore --source HEAD~2 config.yaml # example: go back 2 commits for a file git restore --source origin/main <file> # restore to remote state</code></pre> </a> Rebase</h3> Rebase replays commits on top of another base — producing a linear history. Preferred over merge in feature-branch workflows. H---I---J topicB / E---F---G topicA / A---B---C---D master git rebase --onto master topicA topicB H'--I'--J' topicB / | E---F---G topicA |/ A---B---C---D master</code></pre>git rebase --onto <source-branch> <topic-branch> # play topic commit one-by-one onto source # remove conflicts git rebase --continue git rebase --abort ### Interactive git rebase -i HEAD~3 # range of HEAD</code></pre></a> Reorder</h3> git rebase -i HEAD~2 # reorder as you need & Be careful when performing this step. Accidentally deleting a `commit line` will cause Git to destroy the commit. # line order only changes below pick 7f9d4bf Updated README -> pick 3f8e810 Accessibility fix pick 3f8e810 Accessibility fix -> pick 7f9d4bf Updated README</code></pre></a> Pick</h3> # p, pick <commit> = use commit # nothing changes `pick` 7f9d4bf E:Accessibility fix -> `pick` 7f9d4bf E:Accessibility fix `pick` 3f8e810 F:Updated screenreader -> `pick` 3f8e810 F:Updated screenreader</code></pre></a> Squash</h3> # s, squash <commit> = use commit, but meld into previous commit pick 7f9d4bf E:Accessibility fix -> pick 7f9d4bf E:Accessibility fix `pick` 3f8e810 F:Updated screenreader -> `squash` 3f8e810 F:Updated screenreader `pick` ec48d74 G:Added comments -> `squash` ec48d74 G:Added comments </code></pre>A <- B <- C <- D <- E <- F <- G git rebase -i HEAD~3 A <- B <- C <- D ^-- E={E+F+G}</code></pre></a> Split</h3> # splitting # use commit, but stop for amending pick 7f9d4bf E:Accessibility fix -> pick 7f9d4bf E:Accessibility fix `pick` 3f8e810 F:Updated screenreader -> `edit` 3f8e810 F:Updated screenreader `pick` ec48d74 G:Added comments -> `pick` ec48d74 G:Added comments Stopped at 3f8e810... F:Updated screenreader # OUR Untracked File # untrack1.cs untrack5.cs untrack4.cs untrack3.cs untrack2.cs git add untrack1.cs untrack2.cs untrack3.cs untrack4.cs && git commit -m "Added untrack-files" git add untrack5.cs && git commit -m "Fixed bug untrack-files" git rebase --continue $ git log --oneline --no-pager A <- B <- C <- D <- E <- F <- "Added untrack-files" <- "Fixed bug untrack-files" <- "G" </code></pre></a> Reword</h3> Note: # most recent commit's message use --amend</code> git rebase -i HEAD~2 pick 7f9d4bf Frontpag bug fiz -> reword 7f9d4bf Frontpag bug fix pick 3f8e810 Refactored navbar -> pick 3f8e810 Refactored navbar # Commit message</code></pre></a> Delete</h3> # d, drop <commit> = remove commit git rebase -i HEAD~3 # delete it pick 2f8e823 Refactored -> pick 2f8e823 Refactored pick 7f9d4bf Updated README -> drop 7f9d4bf Updated README pick 3f8e810 Accessibility fix -> drop 3f8e810 Accessibility fix</code></pre> </a> Cherry pick</h3> git cherry-pick -e <commit-hash> # Cherry-pick and edit commit message git cherry-pick <commit-hash> # Cherry-pick single commit # Cherry-pick multiple commits git cherry-pick <commit1> <commit2> git cherry-pick <commit1>..<commit2> </code></pre> </a> Reflog</h3> it records every movement of HEAD, even across resets, rebases, and deleted branches. Entries expire after 90 days by default. </a> Bisect</h3> Helps to find the exact commit that introduced a bug </a> Submodules</h3> Include another repository inside your project using git submodule or git subtree. </a> Advance commands</h3> git gc # Garbage collection git fsck # Verify repository integrity git repack -ad # Optimize repository git prune # Prune unreachable objects git count-objects -vH # Show repository size git blame file.txt # Show who changed each line of file git clean -n # Show what would be removed git clean -f # Remove untracked files git clean -fd # Remove untracked files and directories git clean -fdx # Remove ignored files as well git clean -i # Interactive clean</code></pre> </a> NOTES::</h5> git reset –hard deleted uncommitted work</li> </ol> `git reset --hard` discards both staged and unstaged changes permanently — no undo. :( 1. `git fsck --lost-found --full --no-reflogs --unreachable` 2. cd `.git/lost-found/` Checking object directories: 100% (256/256), done. - unreachable blob `4413d81d94e76ea9a7f7741786f63426d71d297a` - unreachable blob `e613e50292dd773b7e829f598d58e6d419301ec1` - unreachable tree 4b825dc642cb6eb9a060e54bf8d69288fbee4904 git show --stat `4413d81d94e76ea9a7f7741786f63426d71d297a` > recovery-file-a.txt git show --stat `e613e50292dd773b7e829f598d58e6d419301ec1` > recovery-file-a.txt</code></pre> </li> </ol> Running `git checkout <commit-hash>` moves HEAD directly to a commit instead of a branch, so any new commits won't belong to any branch. before-leaving `git checkout -b new-branch-name`, to keep work done in detached HEAD</code></pre> Pushed .env file to sensitive data</li> </ol> revoke and rotate exposed credentials `git filter-branch --force --index-filter 'git rm --cached --ignore-unmatch .env' HEAD` clean the history`git push origin --force --all`</code></pre> Push rejected — non-fast-forward</li> </ol> Never do `git push --force` overrites others work. Use `--force-with-lease` at minimum. 1. pull `git pull origin main` 2. resolve `merge conflict` 3. push `git push origin main` 1. rebase it `git pull --rebase origin main` To globally configure use, `git config --global pull.rebase true`</code></pre> Wrong commit message</li> </ol> 1. `git commit --amend -m 'Correct message here'` 2. if already pushed then do `git push --force-with-lease`</code></pre> Accidentally delete branch</li> </ol> look in `git reflog` for the last commit msg on the deleted branch and recreate using `git checkout -b recovered-branch <sha>`</code></pre> Forget to add file in the last commit message</li> </ol> `git add forgetten-file` and do `git commit --amend --no-edit`. If it's pushed then `git push --force-with-lease`</code></pre> Need to undo a commit that’s already been pushed and shared</li> </ol> No reset, do revert.</code></pre> Identify the large Object</li> </ol> git rev-list --objects --all | sort -k2 | tail -20</code></pre> Wrong files are being tracked</li> </ol> git rm -r --cached node_modules/</code></pre> Empty commit</li> </ol> git commit --allow-empty -m "chore: retrigger CI build"</code></pre> Skip CI</li> </ol> git commit --no-verify #Git hook bypass </code></pre> Bash Scriting - Intermediate 2023-05-01T15:01:35+00:00 Table Of Content</summary> </details> </a> Bash - the REPL Shell</h1> Bash is REPL: [Read Eval Print Loop] You type text, Bash reads it, evaluates it as code, prints the result, then waits for more. This tiny loop quietly runs half the modern internet. </a> Terminal Basics</h2> The terminal is your cmd to the operating system. These commands let you look around, move through directories, and create, inspect, or modify files without a GUI. Mastering these primitives turns the shell from a typing exercise into a precise navigation and file-manipulation tool </blockquote> ls -lah # list all files and dir in current folder pwd # write pathfile of current folder touch filename.extension # create file rm -i file # interactively remove from system. os.Remove(file) mkdir -pv dirname # make folder rmdir -pv dirnam # remove folder cd destination_place # change current directory => destination_place cat -n filename # print it in console head -n num_of_lines filename tail -n num_of_lines filename echo hello world > file.txt # redirect/overwrite in the file echo "bye" >> file.txt # append file</code></pre> </a> Searching Text with grep</code></h2> grep</code> searches text using patterns, not just plain words. It scans line by line and prints only the lines that match. This makes it one of the most powerful tools for exploring logs, configs, and source code—especially when combined with regular expressions </blockquote> grep -A1 pattern file # 1 line after grep -B1 pattern file # 1 line before grep -C2 pattern file # Context ( 2 Line ) grep -i pattern file # ignore case-sensitive grep -o '^.' file # starts with single line grep '^starts_with' file grep 'ends_with$' file grep 'regex-pattern' filename grep -o '^dog' filename # first starts with ^cat of each line cat README.md | grep -i "pattern"</code></pre> </a> paging the file</h2> Paging tools let you view large files one screen at a time instead of dumping everything into the terminal at once. They’re essential when reading logs, manuals, or command output that doesn’t fit on a single screen. </blockquote> less filename # less is more :) more filename</code></pre> </a> Command Introspection</h2> Command introspection helps you answer a deceptively important question: what exactly is Bash running when you type a command? A name might resolve to a builtin, an alias, a keyword, or an external binary—and knowing which one you’re invoking prevents subtle surprises. </blockquote> compgen -b > builtins_cmd.txt # list all building commands compgen -a # alias compgen -k # bash keywords which cmd type -a cmd # shows builtin/alias/hashed/bin-address file </code></pre> </a> Environment & Identity</h2> The shell runs inside a context </blockquote> # environment path echo $PATH | tr : '\n' # translate old new echo $PWD echo $MACHTYPE echo -n $HOSTNAME # no new-line at end whoami uname -a</code></pre> </a> Variable & Command Subsitutions</h2> Variables store text values, and command substitution lets you capture the output of a command as a value. </blockquote> # init variable AWS_S3="sk-12345" # Always DOUBLE QUOTE the variable echo "$AWS_S3" AWS_USER=`uname -a` # Old School way of command substitution AWS_USER=$(uname -m) # Best Pratice of command substitution/ preferred echo "$AWS_USER, and password $AWS_S3"</code></pre> </a> File Permission</h2> Every file in Unix has an access contract: who can read it, who can modify it, and who can execute it. File permissions enforce that contract, and the shell gives you direct control over it. </blockquote> # File Permission # permission ? owned-by group file-size modified-date dir/file file filename chmod +x file # change permission of the file to executable mode. chmod 755 file</code></pre> </a> Shebang & Exit Code</h2> #!/usr/bin/bash echo $? # check any error in the last cmd run</code></pre>SHELL = /usr/bin/bash .SHELLFLAGS := -eu -o pipefail -c PYTHON ?= $(shell command -v python3 || command -v python) PIP = $(PYTHON) -m pip</code></pre> </a> Input</h2> read first_name last_name > Linux love bash read -r variable_address > voc nagar read -rp "Enter your linux OS: " name > Arch read -rsp "Password: " os_password > echo "$first_name, $last_name" # Linux, love bash echo "$variable_address" # voc nagar echo "$name" # Arch echo "$os_password" # </code></pre># Postional Params as Input # $0 # Script name # $1 # first args # $2 # seconda arg # $@ # all args (safe) # $ # all args (unsafe) # $# # args count</code></pre> </a> Conditional</h2> #!/usr/bin/bash if [[ -n $1 ]]; then #help '[[' #$1 <Second argment while calling script> echo "Hi, $1" else # read only one line read -rp "Enter your name: " stdin # -r means raw-string; -p prompt echo "hello, $stdin" fi help '[' # test help '[[' help let # [[ -a ]] True if file exists # [[ -d FILE ]] True if file is a directory. # [[ -e FILE ]] True if file exists. # [[ -f FILE ]] True if file exists and is a regular file. # [[ FILE1 -nt FILE2 ]] True if file1 is newer than file2 (according to modification date). # [[ FILE1 -ot FILE2 ]] True if file1 is older than file2 # [[ -z STRING ]] True if string is empty. # [[ -n STRING ]] True if string is not empty. # [[ STRING1 = STRING2 ]] True if the strings are equal. # [[ STRING1 != STRING2 ]] True if the strings are not equal. # [[ STRING1 < STRING2 ]] True if STRING1 sorts before STRING2 lexicographically. # [[ STRING1 > STRING2 ]] True if STRING1 sorts after STRING2 lexicographically. # [[ ! EXPR ]] True if expr is false. # [[ EXPR1 -a EXPR2 ]] True if both expr1 AND expr2 are true. # [[ EXPR1 -o EXPR2 ]] True if either expr1 OR expr2 is true.</code></pre> </a> Tests & Comparisons</h2> #!/bin/bash a=2 b=2 if [[ $a == $b ]]; then echo 'a and b are same' fi c=3 d=4 if [[ $c != $d ]]; then echo 'c and d are not same' fi if ! [[ $c ==$d ]]; then echo 'C and D are not same' fi if [[ -f file.txt ]]; then echo "file.txt exists and is a file" else echo "file.txt not exists" fi</code></pre>true # $? => 0 false # $? => 1</code></pre> </a> Case statement</h2> #!/usr/bin/bash s=$1 case "$s" in girl) echo Hello, Girl ;; boy) echo Hello, boy ;; kamal* | rev) echo there, you are ;; lo) echo anything starts with love ;& # fall-through love) echo love is all ;; ) echo "Hello $1" ;; esac</code></pre> </a> Control Flow & Loops</h2> #!/usr/bin/bash for i in "0", "muthu", kamalan; do # For loop echo "i: $i" done for i in "$1","$2","$3","$@"; do # For loop echo "i: $i" done while [[ -f file.txt ]]; do echo "file.txt exists" sleep 1 done for i in {a..f};do # [a,b,c,d,e,f] echo "count, $i" done for i in {1..5};do # [1,2,3,4,5] echo "count, $i" done max=5 for ((i=0;i<max; i++)); do echo "count, $i" done</code></pre> </a> Function & Scope</h2> #!/usr/bin/bash greet(){ local name=$1 # by-default, global variable echo "hello, $name from function" } for name in "$@"; do greet "$name" done</code></pre>#!/usr/bin/bash hello(){ local s=$1 echo "hello, $s!" return 0 } goodbye(){ local s=$1 echo "goodbye, $s!" > &1 } # IO Redirections:: stdin,stdout, stderr cat missingfile 2> error.txt # to redirect error msg cat missingfile 2> /dev/null # throw to black whole # help '((' # $# tells number of args # &0 stdin # &1 stdout # &2 stderr if (($#==0)); then echo 'name required! >&2' exit 1 fi for name in "$@"; do if [[ $name == m ]]; then hello "$name" elif [[ $name == k ]]; then hello "$name" else goodbye "$name" fi done</code></pre> </a> Reading Files Safely</h2> #!/usr/bin/bash read -rp "Enter your name: " word echo "single word: $word" # sometime use `echo -n ???` new line won't populate so we use `[[ -n $line ]]` while read -r line || [[ -n $line ]]; do # read from file line by line : done help : # Null command</code></pre> </a> Indexed Array</h2> #!/bin/bash declare -a array=( muthu revathi "voc nagar" kovilpatti ) echo "${array[0]}" echo "${array[1]}" echo "${array[2]}" echo "${array[-1]}" echo "${array[-2]}" echo "${array[2000000]}" # returns "\n" idx=2 echo "idx: ${array[idx]}" # valid echo "idx: ${array[$idx]}" echo "using : ${array[]}" # stringfy echo "using @ : ${array[@]}" for item in "${array[]}"; do echo "item in * is: $item" done for item in "${array[@]}"; do echo "item in @ is: $item" done members="${array[@]}" # stringfy and copy it # WRONG WAY to do it family=( "${array[@]}" mariappan lakshmi ) family+=(rajapaul priya) for m in "${family[@]}"; do echo "item is: $m" done declare -p family alliswell=([0]="atom" [27]="muthu" [24]="revathi") echo "${alliswell[@]}" echo "${#alliswell[@]}" # length of an array</code></pre> </a> Associative Array </h2> #!/usr/bin/bash if ! declare -A fam; then echo "not able to make it" >&2 exit 1 fi declare -A fam fam[husband]="kamalan" fam[wife]="revathi" echo "${fam[husband]}" echo "${fam[wife]}" echo "values: ${fam[@]}" echo "keys: ${!fam[@]}" echo "stringfy: ${fam[]}" echo "length: ${#fam[@]}" for key in "${!fam[@]}"; do val=${fam[$key]} echo "key, $key got value $val" done</code></pre> </a> IFS variable</h2> Internal Field Separator, and it tells Bash how to split text into pieces. </blockquote> #!/bin/bash declare -a fam=( "kamalan" "revathi" "voc nagar" ) echo "stringfy: ${fam[]}" IFS=, echo "stringfy with ",": ${fam[]}" ( # sub shell / local scope IFS=\" echo "stringfy with \": ${fam[]}" ) ( # sub shell / locl scope IFS=! echo "stringfy with !: ${fam[]}" ) echo "stringfy: ${fam[]}"</code></pre> </a> command subsitituion vs Same Shell</h2> #!/bin/bash i=100 say_hi(){ i=6 # local scope echo "hi, max-score $i" } hello=$(say_hi) # runs in sub-shell echo "Greeting and $hello, there" echo "i is $i" # global scope</code></pre>#!/bin/bash i=100 say_hi(){ i=6 # local scope echo "hi, max-score $i" } say_hi # runs in same-shell echo "Greeting and $hello, there" echo "i is $i" # global scope</code></pre> </a> Arithmetic expression</h2> #!/usr/bin/bash a=100000 b=100 echo $(( $a$a )) # EXPRESSION (( max = a > b ? a : b )) echo "$max" echo "$(( "word"2 ))" # => 0 help let a=010 #Octa echo "$a" echo $(( 10#$a ))</code></pre> </a> process subsitution</h2> #!/usr/bin/bash i=0 while read -r word; do echo "$word" ((i++)) done < <(grep d /usr/file.txt) ( # NOTE: `<(grep d /usr/file.txt)` is a file # NOTE: `read` #read it in streaming mode # File descriptor: A file descriptor in Linux (or Unix-like systems) is a small non-negative integer that a process uses as a handle to access an open file or I/O resource. :: <()</code></pre></a> file descriptor in C program</h3> #include <unistd.h> // read, close #include <fcntl.h> // open #include <stdio.h> // perror int main() { char buffer[10]; int fd = open("LICENsSE", O_RDONLY); printf("File descriptor returned by open(): %d\n", fd); if (fd < 0) { perror("open"); return 1; } int bytes_read = read(fd, buffer, sizeof(buffer)); if (bytes_read < 0) { perror("read"); close(fd); return 1; } write(1, buffer, bytes_read); // write to stdout (fd 1) close(fd); write(1, "\n", 1); return 0; }</code></pre>#!/usr/bin/bash # NOTE: shift # means move 2nd argument to 1st $2->$1</code></pre> </a> Text Processing: cut, tr, awk, sed</h2> #!/usr/bin/bash cat <<EOF > file.csv name,age,city,capital kamalan,27,kvp,chennai revathi,24,cvp,"chennai, TN" EOF cat file.csv | tr , "\t" # translate , => \t cat file.csv | cut -d , -f 1-2,4- # cut it by delimeter , and get field of column 1 <index by 1></code></pre>#!/usr/bin/bash man 5 passwd # • login name # • optional encrypted password # • numerical user ID # • numerical group ID # • user name or comment field # • user home directory # • optional user command interpreter cat /etc/passwd | cut -d : -f 1,3,4 # print only login name # TODO cat /etc/passwd | awk -F: '{ print $1, $7 }' cat /etc/passwd | awk -F: '{ printf("name: %s\n",$1 ); }' cat /etc/passwd | awk -F: ' $1=="root" { printf("name: %s\n",$1 ); }' cat /etc/passwd | awk -F: '{ printf("%s\n",$1 ); }' | sort | uniq | wc -l wc -l # lines wc -w # words wc -c</code></pre>#!/usr/bin/bash declare -A shells while IFS=: read -r name pass uid gid gecos home shell; do # if [[ $name == '#' ]]; then continue fi shells[$name]=$shell echo "$name has the shell: ${shells[$name]}" done < /etc/passwd declare -p shells</code></pre> </a> Finding Files</h2> Note:</code> Use find with -exec for bulk operations find location -type f # check only file find location -type f -name '.md' # check only file find location -type d # check only file find location -type l # check only symlinks find location -type f -name '.md' -exec echo I found a file {} \; # execute will run cmd on streaming/one-by-one find location -type f -name '.md' -exec echo I found a file {} ';' find location -type f -name '.md' | xargs # batch it find location -type f -name '.md' -print0 | xargs -0 # separate null to new lines locate --help</code></pre> </a> Debug mode</h2> PS4='[debug]: ' bash -x script.sh bash -n script.sh # syntax check bash -u script.sh # check for undefined variable sellcheck script.sh</code></pre>#!/usr/bin/bash set -x .... # part of the section runs on debug mode .... set +x</code></pre> </a> Pipestatus</h2> #!usr/bin/bash cat file_not_exists.txt | tr , "\t" echo "${PIPESTATUS[]}" true| false | true| false echo "${PIPESTATUS[]}" # Each command in a pipeline gets its own exit code.</code></pre> </a> Timing command</h2> time <---> real # wall clock user # how much time it waits in user space sys # how much time it waits in sys space</code></pre> </a> Sourcing File</h2> # Create and populate into a file cat <<EOF > /tmp/greeetings.sh #!/bin/usr/bash greet(){ # run in local so don't mess with global scope local name=$1 echo "hi, $name" } bye()( # runs in sub-shell name=$1 echo "bye, $name" ) if ! (return 2>/dev/null); then # we are being called directly similar to if name=="main": greet test1 bye test2 EOF cat <<EOF > /tmp/test.sh #!/usr/bin/bash source /tmp/greeetings.sh || exit 1 greet muthukamalan bye revathi EOF chmod +x /tmp/greeetings.sh chmod +x /tmp/test.sh bash /tmp/test.sh</code></pre> </a> params expansion</h2> name="mutHu kamaLan" echo "$name" echo "name: ${name}" echo "substring: ${name:0: -3}" # :- will take precedence as default value subsitution echo "name with capital letter: ${name^}" echo "name with caps all: ${name^^}" echo "name with caps all: ${name@U}" echo "name with cap 1st A: ${name^a}" echo "name with cap all A:${name^^a}" echo "name with cap all M,K: ${name^^[mk]}" echo "${name,}" echo "${name,,[mk]}" echo "replace a with u: ${name/u/a}" echo "replace a with u: ${name//u/a}" echo "replace a with uh: ${name//[uh]/a}" echo "replace with ${name//[uh]/&}" echo "replace with & ${name//[uh]/\&}" path=$(pwd) dirname $path basename $path echo "${path}" echo "${path#/}" # from the beginning echo "${path##/}" # from the beginning but find last occurance echo "${path%/}" # from the end pick 1st / occurance <print "starting to /"> echo "${path%/}"</code></pre>#!/usr/bin/bash name=${1:-default_val} # default value echo "name: ${name}" echo -e "hello,\t${name}" echo -e "hello,\v${name}" echo -e "hello,\b${name}" useros=${1:-$USER} # default assing if not passing echo "hello, $useros!" useros=${1:?'You must pass name'} # show error if not passing echo "hello, $useros!"</code></pre>#!/usr/bin/bash s="muthukamalan" len=${#s} for ((i=0;i<len;i++)); do c=${s:i:1} echo "$c" done</code></pre> </a> Array Expansion</h2> #!/usr/bin/bash fam=( kamalan revathi "voc nagar" ) printf '%s\n' "${fam[@]}" printf '%s\n' "${fam[@]:2}" printf '%s\n' "${fam[@]1:2}" printf '%s\n' "${fam[@]//'voc'/'cvp'}" (IFS=,; echo "${fam[]}")</code></pre> </a> Globbing</h2> ls -1 ./.md ls -1 ./.?? shopt extglob shopt -s extglob # turn on shopt -u extglob # turn off ls -1 ./!(.md) ls -1 ./*.+(yml|md)</code></pre>#/usr/bin/bash touch {foo,bar,baz}.{jpg,txt} printf "<%s>\n" **/ echo {100..50..2} seq 1 20 seq 10</code></pre> </a> brace Expansion</h2> #/usr/bin/bash printf '%s\n' {cvp,tn,india} printf '%s\n' "{cvp,tn,india}" printf '%s\n' tmp/files/{cvp,tn,india}.{jpg,txt} printf '%s\n' tmp/files/.{jpg,txt} fname="pl-" place=( "$fname."{txt,jpg,mov} ) for i in "${place[@]}"; do echo "$i" done</code></pre> </a> printf</h2> #!/usr/bin/bash # printf supports csndiouxXeEfFgGaA printf '<%-5s> %10s\n' hello, world printf '%*s %5s\n' 100 hello, world printf '%d\n' 100 printf '%05d\n' 100 printf '%05d\n' error #=> error printf -v x 'hello, %s' world! echo $x # "hello, world!"</code></pre> </a> Date</h2> date date +%s # strftime datefmt='%Y-%m-%d %H:%M:%S' printf '%(%Y-%m-%d %H:%M:%S)T\n' printf '%(%Y-%m-%d %H:%M:%S)T name=%2s\n' -1 muthu # -1 represent last word</code></pre> </a> SSH Hardening</h2> # Edit SSH config sudo vim /etc/ssh/sshd_config # Set: PermitRootLogin no, Port 2222 sudo systemctl reload sshd # Key-based auth ssh-keygen -t rsa -b 4096 ssh-copy-id devops@server</code></pre> </a> Trap Signals</h2> trap -l</code></pre>#!/usr/bin/bash cleanup(){ echo cleanup function running exit 0 } trap cleanup exit # On exit signal, cleanup fn will run similar to defer() echo script starting.... echo .... echo script done!</code></pre>#!/usr/bin/bash debuggs(){ echo "[debug] running: $BASH_COMMAND" # similar to "bash -x script" exit 0 } trap debuggs debug echo script starting.... echo .... echo script done!</code></pre> </a> Permissions</h2> #!/usr/bin/bash ********** # * d/- # directory or file # * read write execute rwx # owner # * # group # *** # everyone</code></pre>LC_COLLATE=C ls -alh --group-directories-first --color=auto #list all in the current dir. LC_COLLATE=C set of rule recognizing only the US English alphabet — "a-z" and "A-Z" — as "letters", and sorting by byte value</code></pre> </a> Logs, Monitoring and Troubleshooting</h2> top htop uptime w lscpu # system's CPU in use # Architecture: x86_64 # Socket(s): 1 sockets are physical slots on motherboard where we can insert physical cpu. Socket enables communication between two process # Core(s) per socket: 8 each socket will have 8 cores # Thread(s) per core: 2 each core will run 2 threads at the same time # So, Total number of virtual CPU = Sockets x Cores x Threads # CPU(s): 16 lsmem --summary free -g lshw # List Hardware du -lsh filename # size of the file du -sh /var/ # disk usage</code></pre></a> File compression</h2> tar -cf file.tar f1 f2 f3 # .tar called as tarballs -c:create -f:name-of-file tar -tf file.tar # see content of tarball tar -xf file.tar # extract content tar -zcf file.tar f1 f2 f3 # compress tarball to reduce size</code></pre> </a> Check Service and manage daemin</h2> systemctl --help systemctl list-units --all --type= # service, socket path systemctl list-dependencies if systemctl is-active --quiet nginx; then echo "Nginx is running"; fi # Service logs dmesg | tail # Kernel messages journalctl -xb # Boot logs journalctl u nginx journalctl # use journalctl to view and follow the system's journald log entries, which resides in /run/log/journal. runlevel # 5:: Boots into graphical interface and 3:: No-GUI systemctl get-default # graphical.target # runlevel 0 -> poweroff.target # runlevel 1 -> rescue.target # runlevel 2 -> multi-user.target # runlevel 3 -> multi-user.target # runlevel 4 -> multi-user.target # runlevel 5 -> graphical.target # runlevel 6 -> reboot.target # Run a process in background nohup python3 app.py ./script.sh & # Kill process kill 1234 # By PID killall python3 # By name pkill -f "app.py" # By pattern </code></pre> </a> Network Essentials</h2> hostname # managed under /etc/hostname hostnamectl # /etc/hosts ping -c4 google.com # ping utility helps for connectivity checking ip route # info ip ss -tulp netstat -tunpl lsof lsof -i :80 # find process y port lspci # list all pci cards like ethernet card, video cards (peripheral component interconnect) lsblk # list all physical disks space <sda> and partitions are sda1,sda2,sda3.. # Trace route traceroute google.com mtr --report google.com # Real-time # View network interfaces ip addr show ip link show</code></pre> </a> Processes</h2> # List the current active process with their statuses, numbers, resource usage, etc. ps auxc # # All processes ps -ef | grep nginx # Filter By Name fg jobs bg #%id</code></pre> </a> CRON</h2> crontab -l crontab -e # seconds(0-59) min(0-59) hour(0-23) day-of-month(1-31) month(1-12) day-of-week(0-7:: sunday=0 or 7) year[optional] cmd # 0 5 * * * /usr/bin/backup.sh # Alternative systemd.timers</code></pre> </a> Service management with SYSTEMD</h2> Systemd Tools Systemctl</li> journelctl</li> </ul> # Systemctl - Mange system state - start/stop/restart/reload - enable/disable - list and manage units - list and update targets</code></pre>journalctl journalctl -b # logs for current boot journalctl -u UNIT # particular unit</code></pre># It should created in /etc/systemd/system/servicename.service [Unit] Description=Postgresql for Py3 Documentation=http://github.com/muthu/postgresql After=postgresql.service [Service] ExecStart= /bin/bash /usr/bin/filename.sh User= Restart=on-failure RestartSec=10 [Install] WantedBy graphical.target #systemctl get-default # starts with following `systemctl start servicename.service` # check the status `systemctl status servicename.service` # Stop the service `systemctl stop servicename.service` # reload services `systemctl daemon-reload` </code></pre> </a> Storage in Linux</h2> # - Disk partitions lsblk fdisk -l /dev/sda df -h # check disk usage # Parition Types:- Primary, Extended and Logical # Parition Schema:- MBR(old), GPT(new:: unlimited partition, no-max size) # gdisk /dev/diskname</code></pre># LOG ROTATE sudo vim /etc/logrotate.d/myapp # Add: /var/log/myapp/*.log { daily rotate 7 compress }</code></pre># Create partition sudo parted /dev/sdb mklabel gpt sudo parted /dev/sdb mkpart primary ext4 0% 100% # Create filesystem sudo mkfs.ext4 -L Data /dev/sdb1 # Mount filesystem sudo mkdir /data sudo mount /dev/sdb1 /data echo "UUID=$(sudo blkid -s UUID -o value /dev/sdb1) /data ext4 defaults 0 2" | sudo tee -a /etc/fstab # Create LVM sudo pvcreate /dev/sdb1 sudo vgcreate vg_data /dev/sdb1 sudo lvcreate -L 10G -n lv_web vg_data sudo mkfs.ext4 /dev/vg_data/lv_web sudo mount /dev/vg_data/lv_web /var/www # Create RAID sudo mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sdb1 /dev/sdc1 sudo mkfs.ext4 /dev/md0 sudo mount /dev/md0 /mnt/raid</code></pre> </a> Create and Managing Users</h2> # create user sudo useradd -m -s /bin/bash devops # With home dir and shell sudo passwd devops # Set password # Create group sudo groupadd developers # Add user to group sudo usermod -aG developers devops # Modify user sudo usermod -L devops # Lock account sudo usermod -U devops # Unlock account sudo usermod -c "DevOps User" devops # Set comment # Delete user sudo userdel -r devops # Remove user and home dir ##### # Set sudo privileges sudo visudo # Add: devops ALL=(ALL) NOPASSWD:ALL # Passwordless sudo # View users and groups id `user` # User and group info getent passwd # List all users getent group # Group all members `getent group devlops` -> developers group sudo cp .bashrc /etc/skel/.bashrc # customize new user home dir</code></pre> UseFul Extensions 2023-03-01T05:20:35+00:00 Table Of Content</summary> </details> </a> Installation</h2> Unzip the downloaded file.</li> Open chrome://extensions</code> in Chrome.</li> Enable Developer mode (toggle in the top-right).</li> Click Load unpacked and select the unzipped folder.</li> The extension starts working immediately</li> </ol> </a> Demo</h2> List Linkers</a> </li> TabGroups</a> </li> </ol> Secrets Of Life 2022-12-12T12:01:35+00:00 எங்கோ ஒரு மோட்டு வளர்ந்து மலர்ந்து படர்ந்து பரந்து தேன்றல் தான் விசிட மேகம் தான் தாண்டிட வாசமரியா கானகத்தில் தேசமரியா விழுந்தது காலம் சில கடந்தது கனவுகள் சில மறந்தது தேன்றல் தான் வந்திடுமோ தெய்வம் தான் காத்திடுமோ என தேம்மாத நாள் இல்லை ஓடதா ஒரு நதி ஒன்றாய் ஆக காணாத அப்பறவைகள் களித்தன கொண்டாட்டத்தில் தீராது இந்நாள் தினம் தினம் வந்திடவே கூராத அதன் காதல் மெய்யாய் ஆகும்! எப்படி பறப்பது என்று பறவையிடம் கேட்டேன். உட்கார்ந்து சொல்ல தெரியாது என்னுடன் பற என்றது. எப்படி தொடர்ந்து பறப்பது என்று கேட்டேன், இப்படி தான் என்றது 🕊️✨ Telecom Churn Prediction MLOps 2022-01-09T05:20:35+00:00 Table Of Content</summary> </details> Building an End-to-End Telecom Churn Prediction Pipeline: DVC, Hydra, Optuna, and MLflow in Docker

My Love ❤️

Tamil Tokenizer

MediaToAscii

Git in Practical - Intermediate

Bash Scriting - Intermediate

UseFul Extensions

List Linkers</a> </p> </li>

TabGroups</a> </p> </li> </ol>

Secrets Of Life

Telecom Churn Prediction MLOps

Muthukamalan 👋

LLM Inference

HF Model Memory Usage</a></li>
GPU Poor vRAM Calculator</a></li>
LLM Model VRAM Calculator</a> only for quantization models</li>
LLM Explorer</a> to check raw model vRAM size consumption</li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

VRAM Calculator</a></li>
</a></li>
</a></li> </ul>

k8S ❤️ KServe

Muthukamalan 👋

LLM Inference

k8S ❤️ KServe

Tamil Tokenizer

MediaToAscii

Git in Practical - Intermediate

Bash Scriting - Intermediate

</span></a> paging the file</h2> Paging tools let you view large files one screen at a time instead of dumping everything into the terminal at once. They’re essential when reading logs, manuals, or command output that doesn’t fit on a single screen.</p> </blockquote>

</span></a> Environment & Identity</h2> The shell runs inside a context</p> </blockquote>

</span></a> Variable & Command Subsitutions</h2> Variables store text values, and command substitution lets you capture the output of a command as a value.</p> </blockquote>

</span></a> File Permission</h2> Every file in Unix has an access contract: who can read it, who can modify it, and who can execute it. File permissions enforce that contract, and the shell gives you direct control over it.</p> </blockquote>

</span></a> IFS variable</h2> Internal Field Separator, and it tells Bash how to split text into pieces.</p> </blockquote>

</span></a> Service management with SYSTEMD</h2> Systemd Tools</p> Systemctl</li>

UseFul Extensions

</span></a> Installation</h2> Unzip the downloaded file.</li> Open chrome://extensions</code> in Chrome.</li> Enable Developer mode (toggle in the top-right).</li> Click Load unpacked and select the unzipped folder.</li>

Secrets Of Life

Telecom Churn Prediction MLOps