Swapnil Kulkarni | स्वप्निल कुलकर्णी

Passionate Technologist | Cloud Native Leader | Diverse Interests | *own views*

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

The #goal is to be

"catch(Exception e)"

And create value by handling unknown exceptions 😇

#Experience

[Akshay Sonawane @akkias]

Shambolic service & blatant LIES by @BlueDart_ Pune! marked "Consignee Not Available" while I’m at home—a total FRAUD update. Called numerous times, only to get wrong info & false promises of "call backs" that never happened.

@dhlindia @jagograhakjago

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

@PMCPune @SwachhBharatGov @SwachSurvekshan @majhivasundhara @Pmajhivasundhra Very Often. Both sides of service road on Katraj Bypass from Katraj till Wakad

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

The regular culprits are #autos #cab drivers and two wheeler riders driving in wrong sides or creating additional lanes and choking the opposite side. #Pune #PuneTraffic

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

#Bavdhan and #Baner are creating a mini-Bengaluru, you step out for office or any normal daily chore after 9.30am, you can definitely target 10-15/km traffic snarls.

@PuneCityTraffic @PCcityPolice Need some immediate intervention from traffic authorities.

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

@PuneCityTraffic @PCcityPolice पुणे बायपासवरील पाषाण–सूस फ्लायओव्हरखालील अनधिकृत (निशाणी नसलेले) क्रॉसिंग बंद करण्याची नम्र विनंती आहे. या मार्गावर वाहतुकीची प्रचंड वर्दळ असते. कोणी अचानक थांबल्यास पादचाऱ्यांना किंवा वाहनांना निश्चितच अपघाताचा धोका निर्माण होऊ शकतो.

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

@PuneCityTraffic @PCcityPolice A humble request to close the unmarked crossing below Pashan Sus Flyover on the Pune bypass. The stretch has a descent amount of traffic. It will definitely create some harm either to pedestrians or to vehicles if someone stops suddenly.

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

@nitin_gadkari Please travel from Pune -> Nagpur in a MSRTC bus.

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

@CPPuneCity @CP_PCCity Any such initiatives?

[Commissioner of Police, Greater Mumbai @CPMumbaiPolice]

Every voice heard, every concern given priority.

Mumbaikars are welcome at @MumbaiPolice Headquarters every Tuesday at 3:30 PM to share their concerns and grievances directly with me.

No prior appointment is needed, just walk in.

#MeetCPOnTuesday
#MumbaiPolice4All

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

I have never seen any customer service as awful as @GetOneCardIN where a customer who's failing to install the mobile app has been provided a link to service ticket that opens in the mobile app 😂

@OneCardHelp

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

Dear @timesofindia same page printed on both Mondays (17-nov & 24-Nov)

Need someone to proof read

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

📝 Setting up centralized logging with the OpenTelemetry Collector?

The preset/logs preset makes it almost effortless. Instead of 40+ lines of YAML for receivers, processors, exporters, and pipelines, you can enable logs with just 3 lines. Same result. Less boilerplate. Faster start. ✅

In my latest blog, I break down:

- What the logs preset does under the hood
- When to use it (and when not to)
- How to extend it for real-world pipelines

👉 https://t.co/ntFUSCUaCi

hashtag#OpenTelemetry hashtag#Observability hashtag#Logging hashtag#DevOps hashtag#Presets

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

📊 Want to collect metrics with the OpenTelemetry Collector without writing pages of YAML? No manual wiring of Prometheus receivers, batch processors, or exporters, it’s all built in. 🚀

In my latest blog, I cover:

- What the metrics preset sets up under the hood
- When to use it (and when not to)
- How to extend it for real-world monitoring

👉 https://t.co/SCYW7JFdpE

hashtag#OpenTelemetry hashtag#Metrics hashtag#Observability hashtag#DevOps

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

🔎 Distributed tracing with OpenTelemetry doesn’t have to mean pages of YAML.

With the preset/traces preset, you can stand up a full pipeline with just 3 lines. That’s OTLP receiver + batch processor + OTLP exporter, all wired automatically. 🚀

In my latest blog, I break down:

What preset/traces sets up under the hood
When to use it (and when not to)
How to extend it for real-world tracing pipelines

👉 https://t.co/Nn1uHONcsu

hashtag#OpenTelemetry hashtag#Tracing hashtag#Observability hashtag#DevOps

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

☸️ Getting observability right in Kubernetes is tricky:

Logs scattered across nodes
Metrics buried in kubelet
Traces spanning dozens of services

The preset/kubernetes preset in the OpenTelemetry Collector makes it dead simple, one block = logs, metrics, and traces wired up automatically. 🚀

In my latest blog, I break down:

What the kubernetes preset sets up under the hood
When to use it (and when to go custom)
How to extend it for real-world clusters

👉 https://t.co/4giorOCQhK

hashtag#Kubernetes hashtag#OpenTelemetry hashtag#Observability hashtag#DevOps

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

⚡ OpenTelemetry Collector presets are the fastest way to get telemetry flowing. But they’re not the finish line.

✅ Use presets when:

You’re getting started
You need simple logs/metrics/traces quickly
You want safe defaults with minimal YAML

🛠️ Move to custom configs when:

You need advanced scraping, enrichment, or filtering
You’re exporting to multiple backends
You’re building production-grade pipelines

In my latest post, I break down when to stick with presets and when to graduate to full custom configs, plus a clear migration path.

👉 https://t.co/CbeDoHFBiu

hashtag#OpenTelemetry hashtag#Observability hashtag#DevOps hashtag#CloudNative

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

🚀 Configuring the OpenTelemetry Collector can feel like writing YAML novels.

Receivers, processors, exporters, pipelines… one wrong reference and nothing works. Even a “simple” logs pipeline can take 40+ lines.

That’s why Collector presets exist. They’re built-in shortcuts that spin up common telemetry pipelines (logs, metrics, traces, Kubernetes) with a single block of config. One line vs. a whole page of config. Same result. ✅

In my latest post, I break down why presets exist, the challenges they solve, and how they make getting started with OpenTelemetry way less painful.

👉 https://t.co/0p7lzBVsCm

hashtag#Observability hashtag#OpenTelemetry hashtag#DevOps hashtag#CloudNative

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

🔐 Zero-Trust meets AI in Kubernetes 1.34!

With Pod mTLS Certificates and Resource Health Status, Kubernetes now secures communication and ensures GPU reliability.

✅ Encrypted pod-to-pod trust
✅ Real-time GPU health detection
✅ Built-in zero-trust for AI clusters

Explore the full blog with YAML examples 👇

👉 https://t.co/v1MZ4WDwzY

#Kubernetes #AI #Security #GPU #CloudNative #ZeroTrust

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

🔍 Observability meets AI security in Kubernetes 1.34!

With OpenTelemetry tracing now integrated into core features like DRA and Pod Certificates, you can trace:
✅ GPU allocations
✅ Token and certificate events
✅ Secure pod communication paths

In my latest blog, I explore how Kubernetes 1.34 turns tracing into a security enabler for AI pipelines 👇
👉 https://t.co/Pz7TZAmD6f

#Kubernetes #AI #Security #Observability #OpenTelemetry #CloudNative

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

🔒 Kubernetes 1.34 balances AI performance and zero-trust security!

With Dynamic Resource Allocation (DRA), Pod Certificates, and Pod-scoped tokens, your GPU workloads can now run faster and safer.

✅ Secure GPU isolation
✅ Pod-to-Pod mTLS
✅ Strong identity & token protection

Check out my latest blog for YAML examples and real-world AI security use cases 👇
👉 https://t.co/Tx5GaakY1E

#Kubernetes #AI #GPU #Security #ZeroTrust #CloudNative #Kubernetes134

[Swapnil Kulkarni | स्वप्निल कुलकर्णी @MeSwapnilK]

🚀 Kubernetes 1.34 unlocks smarter GPU scheduling!

With Dynamic Resource Allocation (DRA), Kubernetes can now:
⚡ Dynamically allocate GPUs & accelerators
🔌 Support vendor-specific drivers & policies
📊 Improve scheduling with device health & topology

In my latest blog, I dive into:
•How DRA works
•Example YAML configs
•Advantages & gotchas for AI/ML workloads

👉 Read here: https://t.co/p78pCIKLz8

#Kubernetes #AI #ML #GPU #K8s #Security