BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//HasGeek//NONSGML Funnel//EN
DESCRIPTION:Geeking out on systems and security since 2012
X-WR-CALDESC:Geeking out on systems and security since 2012
NAME:Rootconf Mini 2024 (on 22nd & 23rd Nov)
X-WR-CALNAME:Rootconf Mini 2024 (on 22nd & 23rd Nov)
REFRESH-INTERVAL;VALUE=DURATION:PT12H
SUMMARY:Rootconf Mini 2024 (on 22nd & 23rd Nov)
TIMEZONE-ID:Asia/Kolkata
X-PUBLISHED-TTL:PT12H
X-WR-TIMEZONE:Asia/Kolkata
BEGIN:VEVENT
SUMMARY:Check-in
DTSTART:20241122T033000Z
DTEND:20241122T040000Z
DTSTAMP:20260419T170150Z
UID:session/4V6QoXsGJeThGNFhQ3JewA@hasgeek.com
SEQUENCE:2
CREATED:20240929T133412Z
LAST-MODIFIED:20241108T084413Z
LOCATION:Bangalore
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Check-in in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Introduction to the Systems Engineering Track
DTSTART:20241122T040000Z
DTEND:20241122T041000Z
DTSTAMP:20260419T170150Z
UID:session/HaLtBm2wZDMAVnCm7MaotH@hasgeek.com
SEQUENCE:4
CREATED:20241108T090050Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241110T114303Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Introduction to the Systems Engineering Track in Gallery\, 2nd
  floor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T041000Z
DTEND:20241122T042000Z
DTSTAMP:20260419T170150Z
UID:session/AEs84cgpMMt7KqLvYP1Peg@hasgeek.com
SEQUENCE:3
CREATED:20241108T090646Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241110T114304Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Gallery\, 2nd floor (Systems Engineering Track 1
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Introduction to Security Engineering Track
DTSTART:20241122T041500Z
DTEND:20241122T043000Z
DTSTAMP:20260419T170150Z
UID:session/CDeSSZnA2JnK3LK2sebzkn@hasgeek.com
SEQUENCE:9
CREATED:20240929T133843Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241109T162221Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Introduction to Security Engineering Track in Board Room\, 2nd
  floor (Security Engineering Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Art of caching: ways\, wins\, woes\, weird\, wisdom
DTSTART:20241122T042000Z
DTEND:20241122T045000Z
DTSTAMP:20260419T170150Z
UID:session/RdHSPmVZcvwW4sBxNQEbo@hasgeek.com
SEQUENCE:7
CATEGORIES:40 min talk,Systems engineering
CREATED:20241108T091007Z
DESCRIPTION:## TLDR\;\nAn advanced exploration of war stories from buildin
 g caching systems at a decacorn.\n\n## Description\nThere is a reason why 
 the following quote is famous:\n\n“There are only two hard things in Com
 puter Science: cache invalidation and naming things.” — Phil Karlton\n
 \nCaching is everywhere\; it could be a simple system of storing only the 
 most popular books in a shop and keeping all other books in the backroom\,
  or the sophisticated ways we build caching in software systems. This rang
 es from a simple cache layer on top of a database to using complex data st
 ructures like bloom filters as a cache for more expensive checks to the st
 orage layer.\n\n\n## Topics Covered\n### Ways & Wins\nExploring various fo
 rms of caching systems\, characteristics of a caching system and numerous 
 innovative ways a caching system gets deployed at scale.\n\nExamples: Mult
 iple cache layers\, efficiency gains\, eviction policies\, metrics\, consi
 stency\, Saving on latency\, saving on cost\, serving as a fallback\, pred
 ictable response times\, improved resiliency.\n\n### Woes & Weird\nDespite
  the benefits\, caching systems can introduce complexities and challenges 
 that require careful consideration. They also surface unusual and unexpect
 ed behaviors that emerge when working with caching systems at scale in ter
 ms of use cases and scale.\n\nExamples: Cache Stampede\, Bi-Modal Behavior
 \, Timeout Propagation Issues\, Cache Poisoning\, Key Collisions\, Caching
  Increasing Response Time\, Infinite Loops on Cache Invalidation\, Negativ
 e Caching\, Self-Immolating Caches\, Cache Misses Triggering Batch Jobs\n\
 n### Wisdom\nBest practices and lessons learned from building and maintain
 ing caching systems.\n\nExamples: Effective Key Management\, Monitoring an
 d Profiling\, Handling Cache Errors Gracefully\, Avoiding Misuse of Cachin
 g Systems\, Data Serialization Strategies\n\n## Target Audience & Prerequi
 sites\nThe talk targets slightly experienced folks (1–2+ years of experi
 ence in backend) or even those who are early in their career if they have 
 spent some time in systems thinking.\n\nIn terms of prerequisites\, AWS ha
 s built a set of pretty good resources around caching: https://aws.amazon.
 com/caching/. The Systems Design Roadmap also provides a path to start mor
 e research around caching: https://roadmap.sh/system-design.\n\nFor those 
 interested in a more advanced perspective\, this paper is a very good entr
 y point around patterns and behaviors observed when managing large-scale c
 aching systems (both in terms of infrastructure and various use cases): ht
 tps://www.usenix.org/conference/osdi20/presentation/yang\n\n\n## Slide Dec
 k\nDraft: https://docs.google.com/presentation/d/1L9oHgnhioVKZikgdf-v5918P
 XqWhk9uCF2Pg9QejVac/edit?usp=sharing\n\nDraft V1: https://docs.google.com/
 presentation/d/155ppuYIijtpLSCJDA9jk7c1RFTjY1bFq28Z0sqoBpeo/edit?usp=shari
 ng
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074620Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/art-of-caching-ways-wins-wo
 es-weird-wisdom-RdHSPmVZcvwW4sBxNQEbo
BEGIN:VALARM
ACTION:display
DESCRIPTION:Art of caching: ways\, wins\, woes\, weird\, wisdom in Gallery
 \, 2nd floor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Introduction to Systems Engineering Track 2\; housekeeping rules
DTSTART:20241122T042000Z
DTEND:20241122T043000Z
DTSTAMP:20260419T170150Z
UID:session/Gc3yf9uxBwWebpV3YfWTUC@hasgeek.com
SEQUENCE:10
CREATED:20241109T162319Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T050022Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Introduction to Systems Engineering Track 2\; housekeeping rul
 es in Library\, 2nd floor (Systems Engineering Track 2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T043000Z
DTEND:20241122T044000Z
DTSTAMP:20260419T170150Z
UID:session/QszeYknbuaNj21JmfVeHPb@hasgeek.com
SEQUENCE:2
CREATED:20241108T090602Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241111T084304Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Board Room\, 2nd floor (Security Engineering Tra
 ck) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T043000Z
DTEND:20241122T044000Z
DTSTAMP:20260419T170150Z
UID:session/7UdmB9ZaYKBZifNPQ7KzJX@hasgeek.com
SEQUENCE:6
CREATED:20241111T084700Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T050025Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Library\, 2nd floor (Systems Engineering Track 2
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Defence against the dark arts - building security in application d
 evelopment framework
DTSTART:20241122T044000Z
DTEND:20241122T051000Z
DTSTAMP:20260419T170150Z
UID:session/Gem1deueUU3jvBhgKJPGtX@hasgeek.com
SEQUENCE:13
CATEGORIES:40 min talk,Security
CREATED:20241108T040403Z
DESCRIPTION:Startups must move fast. Does this mean compromising on securi
 ty? Everyone will choose security but no startup will have the resources t
 o establish a matured security program from inception. How do you move fas
 t while staying secure even when you have code contributions from interns\
 , software engineers of different experience levels and multi tasking foun
 ders? \n\nThis talk will share the real world experiences gathered while t
 ackling security in an early stage startup. We will focus on application s
 ecurity only and rely on fundamental security controls such as Authenticat
 ion\, Authorization\, Input Validation\, Data Validation\, Multi-Tenancy\,
  Resource Limits\, Auditing and Observability. All of these built as part 
 of the application development framework and used declaratively by enginee
 rs while developing a feature. Examples will be in Go but the patterns dis
 cussed can be applied in any language or framework.\n\nKey take-aways from
  this talk:\n\n- Using Protocol Buffers & gRPC for API spec based developm
 ent\n- Declarative input validation using protocol buffers\n- Declarative 
 authentication & authorization\n- Hooking into API adapters (ConnectRPC) t
 o enforce security controls\n- Hooking into the ORM for multi-tenancy & re
 source limits\n- Security observability using metrics\n\nThis talk is inte
 nded for\n\n- Startup CTO / Head of Engineering\n- Platform / Security / S
 oftware Engineers\n\nWhile the title indicates startup\, this talk should 
 be useful for any security & engineering team who believes in proactive se
 curity controls built within the internal developer platform.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074608Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/paving-the-path-for-secure-
 software-engineering-for-startups-Gem1deueUU3jvBhgKJPGtX
BEGIN:VALARM
ACTION:display
DESCRIPTION:Defence against the dark arts - building security in applicati
 on development framework in Board Room\, 2nd floor (Security Engineering T
 rack) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Shaken\, not stirred - the art of mixing K8s and databases
DTSTART:20241122T044000Z
DTEND:20241122T051000Z
DTSTAMP:20260419T170150Z
UID:session/SRuwsBa4YNg2LABup2awiU@hasgeek.com
SEQUENCE:23
CATEGORIES:40 min talk,Systems engineering
CREATED:20241016T090424Z
DESCRIPTION:### Abstract\nWhen most companies talk about Kubernetes adopti
 on\, they talk about the stateless aspect. However most of them shy away f
 rom Kubernetes when it comes to the stateful part. In this talk we will ex
 plore why Flipkart chose to move to stateful K8s for databases\, the chall
 enges we faced in this journey and the road ahead.\n\nThis talk will cover
  some of the key technical investments Flipkart has made over the last cou
 ple of years and how some of them are shaping up in significantly improvin
 g productivity of our engineering teams.\n\nAll of the content covered in 
 this talk is coming from real world usage of production databases on K8s s
 tateful at a massive scale over the last 3+ years. \n\n### Key Takeaways\n
 - Learn how to evaluate what makes sense for your Organisation's database 
 stack - should you continue to be on VMs or move to K8s.\n- Get real insig
 hts into the many of the *unexpected* challenges K8s poses when you start 
 managing large fleets of data.\n- Learn about Flipkart's Hybrid Cloud (Fli
 pkart Cloud Platform + Google Cloud Platform) and how we view BCP-DR.\n\n#
 ## Audience\n- **DBAs / DBRE Engineers**: Individuals who are also databas
 e wizards and would like to know how their world is quickly evolving.\n- *
 *Engineering Leaders**: Leaders who are responsible for taking decisions o
 n long term technical investments\, specifically on the cost vs availabili
 ty or build vs buy aspects for databases.\n- **Cloud Architects and Engine
 ers**: Individuals who are responsible for designing\, implementing\, and 
 managing cloud-based infrastructure and applications.\n- **Technically Cur
 ious**: Anyone and everyone who is technically curious about K8s or Hybrid
  Cloud or Databases. This talk will attempt to cover a lot of advanced con
 cepts in an easy to understand way.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074559Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/moving-your-databases-to-ku
 bernetes-flipkarts-dbaas-journey-SRuwsBa4YNg2LABup2awiU
BEGIN:VALARM
ACTION:display
DESCRIPTION:Shaken\, not stirred - the art of mixing K8s and databases in 
 Library\, 2nd floor (Systems Engineering Track 2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T045000Z
DTEND:20241122T050000Z
DTSTAMP:20260419T170150Z
UID:session/HowfQfCjaxqUajTELxygFz@hasgeek.com
SEQUENCE:4
CREATED:20241109T162647Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241110T114308Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Gallery\, 2nd floor (Systems Engineering Track 1
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Bridging the gap: building OpenTelemetry pipelines for observabili
 ty in practice
DTSTART:20241122T050000Z
DTEND:20241122T053500Z
DTSTAMP:20260419T170150Z
UID:session/YahdDy6fcBrujH7ynSfkY6@hasgeek.com
SEQUENCE:11
CATEGORIES:40 min talk,Systems engineering
CREATED:20241109T162944Z
DESCRIPTION:\n### Introduction\n\n  The observability stack in modern orga
 nizations often consists of multiple vendors handling logs\, metrics\, and
  traces. This results in inconsistent data formats and conventions\, incre
 asing the operational overhead of maintaining these pipelines. \n\nSREs mu
 st navigate various agents\, each with its debugging methods. Integrated s
 olutions that combine logs\, metrics\, and traces tend to use proprietary 
 formats\, leading to vendor lock-in. These all-in-one solutions often prov
 ide only the most basic features\, creating gaps in production observabili
 ty.\n\nThe OpenTelemetry (OTel) project aims to address these issues. As o
 ne of the largest CNCF projects in the observability space\, it offers a u
 nified agent that uses open standards to collect and send logs\, metrics\,
  and traces. OTel supports multiple observability backends\, reducing the 
 complexity and avoiding vendor lock-in.\n\n### What are the building block
 s of OTel?\n\n  OTel can be overwhelming for beginners due to its many ter
 minologies and steep learning curve. We will learn the essential basics yo
 u need to know to set up your own OpenTelemetry pipeline.\n\n### What does
  it take to run OTel in production?\n\n  OTel's adoption is still evolving
 \, and while there are many resources available\, structured guidance on r
 unning it in production is sometimes hard to find. We will explore how to 
 architect systems handling terabytes of production data through OTel pipel
 ines. Some examples of these systems include:\n\n  - **Datadog agents** pe
 rforming dual writes to both Datadog and OTel-supported backends.\n  - **T
 hird-party solutions** streaming data to AWS S3\, which is then pushed to 
 OTel-supported backends (an approach used by platforms like Cloudflare and
  Fastly).\n  - Leveraging **AWS S3 as cold storage** to rehydrate your obs
 ervability backend when needed.\n  - **Integrating open-source shippers** 
 like Fluentd and Logstash with OTel for seamless data flow.\n\n### Key tak
 eaways for the Audience\n\n- **Clear understanding of OTel building blocks
 **: A straightforward explanation of the essential components of OpenTelem
 etry.\n- **Integrating OTel in production**: Practical guidance on how to 
 implement OTel for various production workloads.\n- **Architectural design
  patterns**:  Design patterns for setting up OpenTelemetry in different sy
 stem architectures.\n- **Migrating from proprietary solutions to OTel**: O
 verview of the approaches taken to migrate from vendor-specific observabil
 ity tools to OpenTelemetry.\n- **Reference setups**: I will share a github
  repository for most of the above OTel configurations.\n
GEO:12.9317;77.6227
LAST-MODIFIED:20251123T071705Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/building-opentelemetry-pipe
 lines-YahdDy6fcBrujH7ynSfkY6
BEGIN:VALARM
ACTION:display
DESCRIPTION:Bridging the gap: building OpenTelemetry pipelines for observa
 bility in practice in Gallery\, 2nd floor (Systems Engineering Track 1) in
  5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T051000Z
DTEND:20241122T052000Z
DTSTAMP:20260419T170150Z
UID:session/TNqYK1D5Une3LrYve1dtBp@hasgeek.com
SEQUENCE:10
CREATED:20240929T133912Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241111T084308Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Board Room\, 2nd floor (Security Engineering Tra
 ck) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T051000Z
DTEND:20241122T052000Z
DTSTAMP:20260419T170150Z
UID:session/58cGyBFQhGtKF83Wu1VFeH@hasgeek.com
SEQUENCE:3
CREATED:20241114T045855Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T050028Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Library\, 2nd floor (Systems Engineering Track 2
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:BOLT-ing CI: securing GitHub actions
DTSTART:20241122T052000Z
DTEND:20241122T053500Z
DTSTAMP:20260419T170150Z
UID:session/2DbmZF3rebxYzehJbnp8u3@hasgeek.com
SEQUENCE:14
CATEGORIES:15 mins demo or experience report,Security
CREATED:20241108T064951Z
DESCRIPTION:**Abstract**:\nCI systems are the security orchestration centr
 e of the SDLC but CI itself has become an attack surface as Solarwinds and
  Codecov attacks have shown.\n\nBOLT: https://github.com/koalalab-inc/bolt
  an OSS tool\, secures CI runtime/build time from two attack vectors: \n1.
  Secret Exfiltration and\n2. Build Tampering.\n\nBOLT instruments an egres
 s gateway and auditd rules. Egress gateway enables BOLT to filter outbound
  traffic. Auditd logs are used to monitor for any kind of source/build tam
 pering.\n\nFurthermore\, egress-filter itself has many other nuances. The 
 outbound traffic from the build system can go to **multi-tenant systems** 
 like GitHub\, DockerHub\, JFrog etc. Ensuring only the trusted tenants are
  being called in egress calls is necessary to ensure security. \n\nBOLT in
 struments **eBPF probes** to do **deep SSL inspection of egress traffic**.
  Applying destination aware rules on top of deep SSL inspection enables BO
 LT to trust tenants across multi-tenant systems like GitHub/DockerHub/JFro
 g.\n\n![image](https://private-user-images.githubusercontent.com/149300820
 /354898979-c115b4d1-d42c-4e72-85a4-b61eeda83371.png?jwt=eyJhbGciOiJIUzI1Ni
 IsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJ
 jb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3Mjg2NDM1MDcsIm5iZiI6MTcyODY0M
 zIwNywicGF0aCI6Ii8xNDkzMDA4MjAvMzU0ODk4OTc5LWMxMTViNGQxLWQ0MmMtNGU3Mi04NWE
 0LWI2MWVlZGE4MzM3MS5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1Bb
 XotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjQxMDExJTJGdXMtZWFzdC0
 xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI0MTAxMVQxMDQwMDdaJlgtQW16L
 UV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT1iMTE1ZjZiMGZiMzEzNzU2ZDI0YmUyYTEwMjk
 zNTMyM2RlZjMwZmVjYWY1OGJmOGViNGVmYTE3YTlhMDQ1Zjg5JlgtQW16LVNpZ25lZEhlYWRlc
 nM9aG9zdCJ9.O9LMPowPApW7bryutaPreIM0esKVWL50EwslUoI8UDQ)\n\n**Takeaways**:
 \n1. Understand threat vectors in CI systems.\n2. Use BOLT to secure Githu
 b actions CI runtime.\n3. Understand nuances of egress-filter at build tim
 e.\n4. Understanding instrumentation of eBPF filtering and how deep SSL in
 spection can be used for other use-cases like minimal token permissions.\n
 \n**Audience**:\n1. Security Engineers: who are keen to understand and imp
 lement advanced CI security measures.\n2. Security Leadership: CISOs\, Sec
 urity Directors\, and Managers looking to gain insights into the nuances o
 f CI/CD security and software supply chain protection.\n3. Cloud Architect
 s/Platform engineers: Professionals designing secure cloud infrastructures
  who need to understand the latest in CI security and eBPF applications.\n
 4. Technically Curious: Anyone with a technical background interested in l
 earning about advanced concepts in CI security\, eBPF\, and software suppl
 y chain protection\, presented in an accessible manner.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074535Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/github-action-ci-security-w
 ith-bolt-2DbmZF3rebxYzehJbnp8u3
BEGIN:VALARM
ACTION:display
DESCRIPTION:BOLT-ing CI: securing GitHub actions in Board Room\, 2nd floor
  (Security Engineering Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:From sidecars to serverless to VMs  - Tecton's journey to the prom
 ised future and back
DTSTART:20241122T052000Z
DTEND:20241122T055000Z
DTSTAMP:20260419T170150Z
UID:session/U4RvNWaGgaCqpM6E9ZPVB7@hasgeek.com
SEQUENCE:19
CATEGORIES:40 min talk,Systems engineering
CREATED:20241016T090617Z
DESCRIPTION:# Overview:\n\nThis tech talk proposes to dive into the evolut
 ion of Tecton’s real-time compute stack\, a journey that started with si
 decar processes\, moved through serverless architecture\, and ultimately m
 atured into a native service deployed on virtual machines (VMs). The sessi
 on will (hopefully) outline the challenges\, lessons learned\, and enginee
 ring decisions made at each stage.\n\nI'd like to have the following rough
  sections:\n\n## Background and Context:\n - Tecton’s Realtime Data Stac
 k: Introduce Tecton and its real-time data processing requirements for mac
 hine learning (ML) and feature serving\, including how tecton executes use
 r-defined post processing code in realtime feature retreival APIs.\n - Ini
 tial Architecture - Sidecar Process: Explain how the compute stack initial
 ly relied on a sidecar model\, what this architecture entailed\, and its a
 dvantages in simplicity and quick iteration during early-stage development
 .\n\n## Challenges with Sidecar Processes:\n- Resource Contention: Describ
 e the limitations of running sidecars alongside primary services\, especia
 lly concerning resource isolation\, security posture\, network latencies\,
  and scaling issues.\n- Operational Complexity: How managing large-scale\,
  sidecar-based microservices introduced operational overhead and complexit
 y.\n- Poor customer experience: Tecton dictated the environment/available 
 libraries. Users couldn't customize this because the environment was baked
  into our service code.\n\n## Moving to Serverless Functions:\n- The Appea
 l of Serverless: Discuss the decision to explore serverless functions (AWS
  Lambda) to handle real-time compute\, reducing the overhead of managing s
 ervers and improving cost efficiency.\n- We  utilized AWS Lambda to execut
 e aforementioned user-defined post processing code.\n- We **_also_** exten
 ded our usage of AWS Lambda to build an API-driven ingestion service.\n- *
 Benefits*: Elastic scaling\, more flexible and user-managed python environ
 ments for their postprocessing code\, and better security posture.\n- *Lim
 itations* Cold starts\, poor performance even with warm starts\, concurren
 cy limits\, debugging complexities\, vendor lock-in. These challenges affe
 cted our real-time SLAs and generally led to poor user experience.\n\n## E
 volution to Native Service on VMs:\n- What did we get wrong about serverle
 ss?\n  -  For customers\, performance is paramount.\n  -  In our case\, wi
 th a strict SLA but diverse workloads\, serverless implementations are a n
 on-started. The variance in latencies even in the happy path is too large.
 \n  -  The cost for lambda functions is actually higher than a lean servic
 e serving the same workload. Execution times were higher on Lambda.\n  -  
 Embedding state in Lambda (using Lambda layers) is high friction. Users ch
 anges take multiple minutes to reflect in their production workload. This 
 is too long.\n- The Shift to Native Services: After outgrowing serverless 
 solutions\, explain the move towards a native service on VMs. This section
  will cover why VMs were chosen over containers or Kubernetes in this spec
 ific case.\n- *Performance Gains*: Far reduced latencies\, better control 
 over resource allocation\, and improved predictability of performance for 
 real-time ML feature serving. All while providing the same (in fact\, bett
 er) user product experience.\n- Operational Improvements: Describe how the
  switch to VMs simplified monitoring\, debugging\, and scaling at the infr
 astructure level.\n- Cloud portablility. With the right abstractions in pl
 ace\, we were able to deploy these services to GCP with very little lift.\
 n\n## Technical Challenges and Lessons Learned:\n- Architectural Trade-off
 s: Key trade-offs between these architectures (sidecars vs. serverless vs.
  VMs)\, specifically as interpreted by us.\n- Product lessons learned\n## 
 Future Directions:\n- Long-term vision\n\n
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074510Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/rebuilding-tectons-realtime
 -compute-stack-twice-U4RvNWaGgaCqpM6E9ZPVB7
BEGIN:VALARM
ACTION:display
DESCRIPTION:From sidecars to serverless to VMs  - Tecton's journey to the 
 promised future and back in Library\, 2nd floor (Systems Engineering Track
  2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Morning break
DTSTART:20241122T053500Z
DTEND:20241122T055500Z
DTSTAMP:20260419T170150Z
UID:session/61ki7DCpRwkcM83rtatdoh@hasgeek.com
SEQUENCE:10
CREATED:20240929T133944Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042256Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Morning break in Gallery\, 2nd floor (Systems Engineering Trac
 k 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Morning break
DTSTART:20241122T053500Z
DTEND:20241122T055500Z
DTSTAMP:20260419T170150Z
UID:session/YH1GSpgaToyVmdP3WfBSGU@hasgeek.com
SEQUENCE:7
CREATED:20241110T023545Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042301Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Morning break in Board Room\, 2nd floor (Security Engineering 
 Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Keynote livestream
DTSTART:20241122T055000Z
DTEND:20241122T064000Z
DTSTAMP:20260419T170150Z
UID:session/GTPavi5xndvZbbJzhqELEo@hasgeek.com
SEQUENCE:12
CREATED:20241111T084956Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241121T121318Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Keynote livestream in Library\, 2nd floor (Systems Engineering
  Track 2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Keynote: Revisiting abstractions for fun and profit
DTSTART:20241122T055500Z
DTEND:20241122T064000Z
DTSTAMP:20260419T170150Z
UID:session/Bqt3QV14FKJ6BtyAxp3nsB@hasgeek.com
SEQUENCE:7
CATEGORIES:40 min talk,Systems engineering
CREATED:20241109T100936Z
DESCRIPTION:Abstractions are great! They help us think without being overw
 helmed by details. But sometimes they can come at the cost of understandin
 g\, where they become a wall that we don't wish to climb. That limits us\,
  because when they change\, a lot of us can't keep up.\n\nIn this talk\, I
  make the following assertions:\n1. In the last 10-15 yrs\, progress of ac
 cessibility (the good part) in tech via frameworks\, languages and best pr
 actices\, has come at the cost of lesser awareness of internals (the bad p
 art).\n2. Every abstraction has a half life\, stemming from the environmen
 t that birthed it\, in the sense that when that environment changes\, so d
 oes the relevance of that abstraction. 2-3 half lives in\, people overly r
 eliant on that abstraction can find themselves _stuck_.\n3. Timing seems r
 ipe for a change in bunch of different abstractions\, thanks to some excit
 ing progress in both hardware & software\, of which I'll share a few examp
 les.\n4. Preparing ourselves with foundational knowledge allows us to be b
 etter prepared for\, and make the most of these changes.\n
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074458Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/revisiting-abstractions-for
 -fun-profit-Bqt3QV14FKJ6BtyAxp3nsB
BEGIN:VALARM
ACTION:display
DESCRIPTION:Keynote: Revisiting abstractions for fun and profit in Gallery
 \, 2nd floor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Keynote livestream
DTSTART:20241122T055500Z
DTEND:20241122T064000Z
DTSTAMP:20260419T170150Z
UID:session/XLm6gkYdAJvQHDfF2ExA5a@hasgeek.com
SEQUENCE:5
CREATED:20241114T042241Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241121T121330Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Keynote livestream in Board Room\, 2nd floor (Security Enginee
 ring Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T064000Z
DTEND:20241122T065000Z
DTSTAMP:20260419T170150Z
UID:session/MGfXGtuRLQLD4S9oTzjFwR@hasgeek.com
SEQUENCE:2
CREATED:20241111T082214Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042403Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Gallery\, 2nd floor (Systems Engineering Track 1
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:With infinite scale comes infinite bill (and bankruptcy)
DTSTART:20241122T064000Z
DTEND:20241122T071000Z
DTSTAMP:20260419T170150Z
UID:session/QtF19z1wgufPdJNgCKyipN@hasgeek.com
SEQUENCE:18
CATEGORIES:40 min talk,Security
CREATED:20241108T065018Z
DESCRIPTION:What can a bored hacker do with $5? They can do one of the bel
 ow - buy a coffee\, subscribe to some video streaming service or make your
  company bleed 10s to 100s of dollars in cloud bills. \n\nWhile the majori
 ty of companies are already on cloud or have a hybrid cloud setup\, one mu
 st understand that cloud environments are different organisms compared to 
 data centers. It can massively scale your infrastructure. It can massively
  scale your cloud bills as well - pushing your company anywhere between hu
 ge expenses to the brink of bankruptcy.\n\nThis research based talk shines
  light on an underrated bug class specific to cloud environments\, often r
 eferred to as “denial of wallet” (DoW) or “exhaustion of wallet”\,
  among different cloud providers and cloud services.\n\nThe talk tries to 
 answer the reasons why this bug class exists\, public examples of these at
 tacks in the past and the different cloud and VPS providers that are vulne
 rable to this bug class. The talk will also highlight possible solutions t
 hat can be implemented on both sides of the shared responsibility model to
  detect and mitigate bug class - showing ways to mitigate damage from an a
 ttacker willing to spend $500 against your infrastructure.\n\nKey Takeaway
 s: The root cause of DoW\, different services from different cloud provide
 rs that are affected by this issue\, possible solutions to detect and miti
 gate this bug class.\n\nAudience Segment this talk is going to benefit: Cl
 oud Architects\, Cloud Engineers & Cloud Security Professionals.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074437Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/with-infinite-scale-comes-i
 nfinite-bill-and-bankruptcy-QtF19z1wgufPdJNgCKyipN
BEGIN:VALARM
ACTION:display
DESCRIPTION:With infinite scale comes infinite bill (and bankruptcy) in Bo
 ard Room\, 2nd floor (Security Engineering Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Millions saved\, lessons learned: the cloud cost optimization blue
 print
DTSTART:20241122T064000Z
DTEND:20241122T071500Z
DTSTAMP:20260419T170150Z
UID:session/DLXbGj1vvouoAXssEiyvA5@hasgeek.com
SEQUENCE:8
CATEGORIES:40 min talk,Systems engineering
CREATED:20241111T085935Z
DESCRIPTION:Optimizing cloud costs isn't just about deploying tools and ap
 plying technical best practices—it's fundamentally tied to a shift in or
 ganizational culture\, a strategic and deep understanding of latest techno
 logy\, and the identification of both quick wins and long-term investment 
 areas.\n\nIn this session\, through the lens of what transpired in Sharech
 at\, a startup that has every-thing going for it\, billion $ plus investme
 nts\, 50% week-over-week user growth and which quickly became one of GCP's
  largest APAC (Asia-pacific) cloud deployments\, we will go over the key i
 nsights learnt in undertaking this massive cost reduction journey where we
  reduced cloud infrastructure costs from an ARR of 150M USD to < 25M fraug
 ht with very interesting organizational \, cultural and hugely complex tec
 hnical challenges.\n\n**Key Takeaways** \n* When scale hits\, cloud cost b
 urns.  Through this talk\, I will share insights on how to avoid/mitigate 
 this. Adopting cultural changes in your organization is critical to create
  long term sustainability for cloud costs. Specifically I will deep dive i
 nto\n   - Balancing between growth/product initiatives vs optimization and
  aligning business leaders \n    - Fostering a culture of ownership\, acco
 untability and incentivization amongst engineers.\n   - Design cost contro
 ls so that cost doesnt claw back.\n\n* As companies focus on trying to ach
 ieve PMF\, a lot of bad Design and Architectural decisions get taken which
  gets exposed when scale increases. In this session I  will share insights
  around the following aspects to help you arrive at the right decisions.\n
   - Cloud pitfalls to avoid\n  - Technologies that are hard to get right a
 nd require heavy investment.\n  - Optimization areas where there are quick
  ROI\n  - Adopting evolutionary architectures\n\n**Audience**\n * **Engine
 ering Leaders**: Captains who can take-back insights from this talk and ge
 t to apply in their ever-growing ship.\n* **Cloud Architects and Engineers
 **: Individuals who are responsible for designing\, and managing cloud-bas
 ed infrastructure and applications.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074448Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/millions-saved-lessons-lear
 ned-the-cloud-cost-optimization-blueprint-DLXbGj1vvouoAXssEiyvA5
BEGIN:VALARM
ACTION:display
DESCRIPTION:Millions saved\, lessons learned: the cloud cost optimization 
 blueprint in Library\, 2nd floor (Systems Engineering Track 2) in 5 minute
 s
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Sponsored talk: Building a seamless hybrid cloud with Kubernetes a
 nd Clutch
DTSTART:20241122T065000Z
DTEND:20241122T071500Z
DTSTAMP:20260419T170150Z
UID:session/FbbU4TJtsZJky6MiERGiKm@hasgeek.com
SEQUENCE:11
CATEGORIES:40 min talk,Systems engineering
CREATED:20241111T082547Z
DESCRIPTION:This talk outlines how we use Clutch as a unified entry point 
 for developers to deploy and monitor workloads across Kubernetes clusters 
 in hybrid cloud environments. \nBy integrating Clutch with ArgoCD\, organi
 zations can streamline multi-cloud operations\, automate migrations\, and 
 proactively manage workload health.\n\nKey topics include:\n\n* Hybrid Clo
 ud Benefits: Enhanced flexibility\, cost savings\, resilience\, and compli
 ance across public and private clouds.\n* Challenges of Multi-Cluster Mana
 gement: Clutch implements guard rails\, including security\, and maintaini
 ng consistent configurations across environments.\n* Unified Entry Point f
 or Deployment and Monitoring: Clutch’s single interface allows developer
 s to deploy\, monitor\, and receive alerts across clusters in AWS\, Google
  Cloud\, Azure\, and on-premises.\n* Automated Workload Migration: Easier 
 and seamless workload migration between K8s clusters in a Hybrid cloud set
 up\n* Enhanced Alerting with Clutch: Clutch alerts inform developer teams 
 of resource utilization spikes\, service health\, and availability. Real-t
 ime notifications enable faster response times\, reduce downtime\, and mai
 ntain application health across clusters.\n\n**Who should attend this talk
 :**\nOrganizations needing to deploy and maintain Kubernetes workloads in 
 hybrid cloud environments\n\n**About the speaker:**\nBharadwaj Embar is a 
 seasoned hybrid cloud infrastructure expert with extensive experience in a
 rchitecting and managing scalable\, resilient systems. He specializes in K
 ubernetes\, Druid\, Temporal\, and the HashiCorp stack\, bringing together
  robust cloud-native technologies to deliver high-performance solutions. W
 ith a background in both development and operations\, Bharadwaj is known f
 or his strategic approach to solving complex infrastructure challenges and
  optimizing data workflows in distributed environments. At Rootconf\, he w
 ill share his insights into the practicalities of cloud integration and th
 e effective use of open-source tools to streamline modern infrastructures.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074427Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/building-a-seamless-hybrid-
 cloud-with-kubernetes-and-clutch-FbbU4TJtsZJky6MiERGiKm
BEGIN:VALARM
ACTION:display
DESCRIPTION:Sponsored talk: Building a seamless hybrid cloud with Kubernet
 es and Clutch in Gallery\, 2nd floor (Systems Engineering Track 1) in 5 mi
 nutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lunch break
DTSTART:20241122T071000Z
DTEND:20241122T081000Z
DTSTAMP:20260419T170150Z
UID:session/5XVBrpyGat2wFT5tjgZWyN@hasgeek.com
SEQUENCE:13
CREATED:20240929T134050Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042650Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Lunch break in Board Room\, 2nd floor (Security Engineering Tr
 ack) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lunch
DTSTART:20241122T071500Z
DTEND:20241122T081500Z
DTSTAMP:20260419T170150Z
UID:session/THLzGD7GNrP1YPtzPRwMAd@hasgeek.com
SEQUENCE:6
CREATED:20241111T082232Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T152043Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Lunch in Gallery\, 2nd floor (Systems Engineering Track 1) in 
 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lunch
DTSTART:20241122T071500Z
DTEND:20241122T075500Z
DTSTAMP:20260419T170150Z
UID:session/KxrAujQWwBKpC5eXR6z4VL@hasgeek.com
SEQUENCE:3
CREATED:20241111T085224Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T043626Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Lunch in Library\, 2nd floor (Systems Engineering Track 2) in 
 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Log Days Are Over: Why we fine-tuned LLMs for LogQL
DTSTART:20241122T075500Z
DTEND:20241122T082500Z
DTSTAMP:20260419T170150Z
UID:session/SBxHpPt2bcGgvBJCMJCBoT@hasgeek.com
SEQUENCE:17
CATEGORIES:40 min talk,Systems engineering
CREATED:20241016T090557Z
DESCRIPTION:## Abstract\nMonitoring and observability tools are a cornerst
 one in debugging processing for any large organization.\n\nThe current sta
 te of log query languages and interfaces in popular log analysis tools pre
 sents significant usability challenges that hinder productivity\, such as 
 a lack of standardized query languages\, steep learning curves\, and high 
 complexity.\n\nLarge Language Models\, while promising for log analysis\, 
 parsing\, and summarization\, haven't quite been explored for log search.\
 n\nThis talk covers our study of creating the first text-to-logql dataset\
 , fine-tuning LLMs for [LogQL](https://grafana.com/docs/loki/latest/query/
 )\, Grafana's Log Query Language\; and evaluating various approaches to us
 ing LLMs to write LogQL queries.\n\nWe have been working on a paper on the
  same topic and are excited to share our learnings with fellow systems ner
 ds!\n\n## Key Takeaways\n- Various challenges to using LLMs to solve log s
 earch problems.\n- Evaluation and performance of various fine-tuned models
  and approaches for LogQL.\n- How to build a dataset for a domain-specific
  language from scratch!\n- How this fits in with a bigger goal of making a
  semantic search engine for observability data.\n\n## Audience\n- Anybody 
 who has written a LogQL/DatadogQL/SPL query recently and used or thought a
 bout using LLMs for it.\n- Systems & Infra engineers interested in LLMs fo
 r observability.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074415Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/chatting-with-logs-an-explo
 ratory-study-on-finetuning-llms-for-logql-SBxHpPt2bcGgvBJCMJCBoT
BEGIN:VALARM
ACTION:display
DESCRIPTION:Log Days Are Over: Why we fine-tuned LLMs for LogQL in Library
 \, 2nd floor (Systems Engineering Track 2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Chatbot security: safeguarding RAG chatbots from induced hostile h
 allucinations
DTSTART:20241122T081000Z
DTEND:20241122T083500Z
DTSTAMP:20260419T170150Z
UID:session/ETEHKpmogDyxv2qoZy4YNq@hasgeek.com
SEQUENCE:20
CATEGORIES:40 min talk,Security
CREATED:20241108T065105Z
DESCRIPTION:Large language models (LLMs) are known to generate unintended 
 inaccurate responses\, often called hallucinations. Most of these are harm
 less mistakes\, like Google AI Overview suggesting to eat a rock a day. Th
 ere's a more concerning possibility: what if an attacker could deliberatel
 y cause specific hallucinations? This could allow the stealthy spread of t
 argeted disinformation.\n\nOur talk will introduce the concept of indirect
  prompt injections and show how malicious documents can be crafted to trig
 ger particular hallucinations when added to a vector database. We'll demon
 strate case studies of proof-of-concept attacks on popular LLM chatbots li
 ke Notion AI\, Microsoft Copilot\, and Slack AI. Finally\, we'll explore s
 ecure design principles and tools to defend against these hidden threats.\
 n\nThis talk is designed for a diverse audience in the AI field. It will b
 e particularly valuable for AI engineers working on LLM applications\, AI 
 security engineers focused on protecting these systems\, product managers 
 overseeing AI-powered products\, and C-level executives making strategic d
 ecisions about AI implementation and security. Whether you're hands-on wit
 h the technology or guiding its use at a high level\, you'll gain crucial 
 insights into this emerging threat and its implications.\n\n## Key takeawa
 ys:\n1. Understand the mechanics of indirect prompt injections and their p
 otential for causing targeted hallucinations in popular LLM applications\n
 2. Learn practical strategies and tools to protect AI systems against mali
 cious hallucinations and enhance overall AI security
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074350Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/malicious-hallucinations-hi
 dden-threats-with-indirect-prompt-injection-ETEHKpmogDyxv2qoZy4YNq
BEGIN:VALARM
ACTION:display
DESCRIPTION:Chatbot security: safeguarding RAG chatbots from induced hosti
 le hallucinations in Board Room\, 2nd floor (Security Engineering Track) i
 n 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Surfing through chaos - running IPL from outside the cricket field
DTSTART:20241122T081500Z
DTEND:20241122T084500Z
DTSTAMP:20260419T170150Z
UID:session/PQe3ixLSuSDG2sbwzuoMDZ@hasgeek.com
SEQUENCE:16
CATEGORIES:40 min talk,Systems engineering
CREATED:20241016T090443Z
DESCRIPTION:Tsunami Traffic\, Traffic Avalanche\, Hockey Stick Curve are s
 ome of the common terms laid out as the benchmark for developing systems a
 t scale.\nAs the first hire for the Jiocinema Devops team @ Viacom18 I got
  the opportunity to work on breaking these benchmarks while maintaining ou
 r Infrastructure on a cluster of Kubernetes clusters \;)\n\n# The match ou
 tside the field\nIn my talk I'll dwell into my experience with planning an
 d execution of the infrastructure for IPL 2023 and 2024. Will highlight ke
 y insights and learnings with special focus on below areas:\n\n1. The prep
 aration for large scale events begin months in advance when we regrously t
 est the infra components and capacity planning to handle the load.\n2. Vis
 ibility on every single component becomes very crucial at scale\, auto dis
 covery helps with that.\n3. Contingency planning is crucial and even a bac
 kup plan should have a backup plan. Your SLAs should be defined and met.\n
 \n# Key Takeaways\n1. Scaling strategies for multiple kubernetes cluster d
 uring high traffic events\n2. Proactive monitoring and alert systems\n3. P
 reemptive planning and failovers mechanisms\n\n# Target Audience\n1. SREs\
 , Devops and Platform engineers\n2. Anyone who loves cricket (everyone\, d
 uh?)
GEO:12.9317;77.6227
LAST-MODIFIED:20260220T072641Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/cricket-match-from-a-devops
 -lens-PQe3ixLSuSDG2sbwzuoMDZ
BEGIN:VALARM
ACTION:display
DESCRIPTION:Surfing through chaos - running IPL from outside the cricket f
 ield in Gallery\, 2nd floor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T082500Z
DTEND:20241122T083500Z
DTSTAMP:20260419T170150Z
UID:session/8PfxETWoPvbnqTzPuSdnr5@hasgeek.com
SEQUENCE:2
CREATED:20241114T043548Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T043630Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Library\, 2nd floor (Systems Engineering Track 2
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Sponsored talk: Using Java in low-latency applications
DTSTART:20241122T083500Z
DTEND:20241122T090000Z
DTSTAMP:20260419T170150Z
UID:session/5i2R8rZ3UGwC9N8zLv6h7m@hasgeek.com
SEQUENCE:14
CATEGORIES:40 min talk,Systems engineering
CREATED:20241110T021638Z
DESCRIPTION:Java is not what we think of immediately when it comes to low-
 latency applications - this is typically the realm of C/C++/Rust\, etc. In
  E6data\, we use Java in many parts of the engine and have successfully us
 ed it in cases where we need  high performance and low latency.\n\nIn this
  talk\, we explore some of the techniques we use in E6data to achieve low 
 latencies.\n\nThese techniques will range from techniques to minimise GC\,
  writing CPU cache efficient code\, threading and concurrency techniques\,
  benchmarking and profiling\, etc. \n\n\nKEY TAKEAWAYS:\n\n1. Get an insig
 ht into some of the techniques for achieving low latency in modern CPU har
 dware\n\n2. Explore how we could implement some of these techniques in Jav
 a\n\n\nAUDIENCE:\n\nDevelopers / enthusiasts working on applications with 
 low latency characteristics or just interested in how would one achieve lo
 w latency.
GEO:12.9317;77.6227
LAST-MODIFIED:20251119T105531Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/using-java-in-low-latency-a
 pplications-5i2R8rZ3UGwC9N8zLv6h7m
BEGIN:VALARM
ACTION:display
DESCRIPTION:Sponsored talk: Using Java in low-latency applications in Libr
 ary\, 2nd floor (Systems Engineering Track 2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T083500Z
DTEND:20241122T084500Z
DTSTAMP:20260419T170150Z
UID:session/76R5Xbeq2o7Q7nbakvxqe9@hasgeek.com
SEQUENCE:4
CREATED:20241108T090144Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042702Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Board Room\, 2nd floor (Security Engineering Tra
 ck) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T084500Z
DTEND:20241122T085500Z
DTSTAMP:20260419T170150Z
UID:session/MA69AChdgwiKXfFhrwcekX@hasgeek.com
SEQUENCE:2
CREATED:20241111T083046Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T043140Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Gallery\, 2nd floor (Systems Engineering Track 1
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Community pitches
DTSTART:20241122T084500Z
DTEND:20241122T090500Z
DTSTAMP:20260419T170150Z
UID:session/9gELcrVHQfsmuJNNs5gyRp@hasgeek.com
SEQUENCE:6
CREATED:20241108T085512Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042723Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Community pitches in Board Room\, 2nd floor (Security Engineer
 ing Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Empowering mobile UI automation at scale: dynamic emulator creatio
 n with PhonePe's Drove infrastructure
DTSTART:20241122T085500Z
DTEND:20241122T092000Z
DTSTAMP:20260419T170150Z
UID:session/8CumJmeUsXLLem9egmAo1M@hasgeek.com
SEQUENCE:11
CATEGORIES:40 min talk,Systems engineering
CREATED:20241111T084014Z
DESCRIPTION:## Abstract\nThis tech talk dives into PhonePe's journey on sc
 aling its UI Automation capabilities by building a comprehensive platform 
 to support its fleet of applications and use cases. Investment in automati
 on was made primarily for two reasons: **Improving Product Quality** and *
 *Improving Org Efficiency**. \n\nThe talk will dive into various challenge
 s faced with scaling UI Automation on Android\, existing VCS systems\, Dev
 ice support and how we built a new platform which uses On-demand Emulator 
 Device Farm using PhonePe Drove Container Orchestration to provide scalabl
 e and cost-effective access to virtual devices for testing purposes.\n\n##
  Key Takeaways\n* Learn how to scale your UI automation infra with the hel
 p of on demand android emulator creation with the help of PhonePe’s Drov
 e infrastructure.\n* Get insights of PhonePe’s Test Automation Platform 
 and how it handles UI automation at scale.\n\n## Audience\n* **Engineering
  Leaders:** Leaders who are responsible for taking decisions on long term 
 technical investments\, specifically on the cost vs availability or build 
 vs buy aspects for databases.\n* **DevOps engineers:** Professionals who a
 re interested in automating and improving security posture in cloud enviro
 nments\n* **Cloud Architects and Engineers:** Individuals who are responsi
 ble for designing\, implementing\, and managing cloud-based infrastructure
  and applications.\n* **Technically Curious:** Anyone and everyone who is 
 technically curious.\n
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074317Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/empowering-mobile-ui-automa
 tion-at-scale-dynamic-emulator-creation-with-phonepes-drove-infrastructure
 -8CumJmeUsXLLem9egmAo1M
BEGIN:VALARM
ACTION:display
DESCRIPTION:Empowering mobile UI automation at scale: dynamic emulator cre
 ation with PhonePe's Drove infrastructure in Gallery\, 2nd floor (Systems 
 Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T090000Z
DTEND:20241122T091000Z
DTSTAMP:20260419T170150Z
UID:session/2mai97njWvQW8TPfBG8pMd@hasgeek.com
SEQUENCE:2
CREATED:20241111T085408Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T043556Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Library\, 2nd floor (Systems Engineering Track 2
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lightning Talks
DTSTART:20241122T090500Z
DTEND:20241122T093000Z
DTSTAMP:20260419T170150Z
UID:session/XhyoCR3herVmWfm9XmXZkE@hasgeek.com
SEQUENCE:10
CREATED:20241110T030951Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042728Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Lightning Talks in Board Room\, 2nd floor (Security Engineerin
 g Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Sponsored Round Table - Running infrastructure with data centres v
 ersus cloud
DTSTART:20241122T091000Z
DTEND:20241122T095500Z
DTSTAMP:20260419T170150Z
UID:session/MptCjrvKhcKu6hvbbZZFWL@hasgeek.com
SEQUENCE:12
CREATED:20241110T021720Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074306Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Sponsored Round Table - Running infrastructure with data centr
 es versus cloud in Library\, 2nd floor (Systems Engineering Track 2) in 5 
 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transiton
DTSTART:20241122T092000Z
DTEND:20241122T093000Z
DTSTAMP:20260419T170150Z
UID:session/EMSY8RfPBFMtnxM4YxyqYC@hasgeek.com
SEQUENCE:3
CREATED:20241111T083116Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T151919Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transiton in Gallery\, 2nd floor (Systems Engineering Track 1)
  in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T093000Z
DTEND:20241122T094000Z
DTSTAMP:20260419T170150Z
UID:session/GbqgjQGqx466KUnVxYVgFv@hasgeek.com
SEQUENCE:1
CREATED:20241114T042754Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T042756Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Board Room\, 2nd floor (Security Engineering Tra
 ck) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:ABC of LLMOps: what it takes to run self-hosted LLMs
DTSTART:20241122T093000Z
DTEND:20241122T101000Z
DTSTAMP:20260419T170150Z
UID:session/UE2cgBZHUJEzFA9usDgdCM@hasgeek.com
SEQUENCE:10
CATEGORIES:40 min talk,Systems engineering
CREATED:20241109T163905Z
DESCRIPTION:LLMs and generative AI have made their way into our day-to-day
  operations. While the wrappers over GPT are a good starting point\, I was
  intrigued by what it takes for an SRE to understand the domain\, identify
  its operational aspects\, and build runbooks around running self-hosted L
 LM models.\n\nCurrently\, too many models are built\, but very few are in 
 production. While many companies are trying to streamline the toolchain\, 
 it is still nascent. The body of work I will discuss is an experiment to b
 uild an understanding of the LLMOps ecosystem.\n\nWe built an internal ser
 ver setup and have explored deploying the models on GPUs instead of relyin
 g on OpenAI.\n\nGoals:\n\n1. Learn the domain from first principles.\n2. B
 uild practices around running models on Kubernetes with GPUs.\n3. Know wha
 t it takes to run and manage Vector databases for storing embeddings.\n4. 
 Use the above knowledge to build and produce RAG applications.\n\nWhat was
  our learning curve:\n\n1. Take the basic concepts of the domain and build
  a mental model of the toolchain and ecosystem.\n2. Explore platforms like
  Ray/Kuberay and model repositories like HuggingFace. Understand their usa
 ge from operational aspects.\n3. Learn basic models like Phi3 and evolve t
 o using advanced models like Llama3.1.\n4. Understand the pipeline from a 
 developer perspective - using frameworks like Langchain\, finding its limi
 tations\, and shifting our codebase to LlamIndex.\n5. Start with toy appli
 cations to explore each tool individually. After gaining a basic understan
 ding\, we moved to building an RAG application for internal usage - we bui
 lt a resume filter application.\n6. Dogfooded it internally and learned mo
 re about prompt engineering\, vector embeddings\, and databases like QDran
 t.\n\nWhat will you gain from this talk:\n\n1. How should you approach thi
 s domain if you are managing backend systems - Our learnings were less fro
 m a pure development view and more from a "How will I run this in producti
 on?" lens.\n2. What does it take to build your home lab\, and where can yo
 u save costs - Is the public cloud cheaper\, or if you are an org thinking
  of investing in the domain\, does buying your hardware make sense?\n3. Th
 e domain of LLMOps is developing\, and we are trying to learn it through e
 xperimentation. You will also gain a perspective and an approach to experi
 menting through this landscape.\n\nThis talk benefits software engineers a
 t all levels but is especially relevant for SRE and DevOps practitioners.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074256Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/abc-of-llmops-what-does-it-
 take-to-run-self-hosted-llms-UE2cgBZHUJEzFA9usDgdCM
BEGIN:VALARM
ACTION:display
DESCRIPTION:ABC of LLMOps: what it takes to run self-hosted LLMs in Galler
 y\, 2nd floor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Security engineering track - wrap-up
DTSTART:20241122T094000Z
DTEND:20241122T100000Z
DTSTAMP:20260419T170150Z
UID:session/1SumvZTjoVaZ8WvFZUUbK@hasgeek.com
SEQUENCE:14
CREATED:20241110T030724Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074237Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Security engineering track - wrap-up in Board Room\, 2nd floor
  (Security Engineering Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Evening break
DTSTART:20241122T095500Z
DTEND:20241122T102500Z
DTSTAMP:20260419T170150Z
UID:session/BzYJhDVbH1Z2doWbyeVnHq@hasgeek.com
SEQUENCE:2
CREATED:20241111T085735Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T043602Z
LOCATION:Library\, 2nd floor (Systems Engineering Track 2) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Evening break in Library\, 2nd floor (Systems Engineering Trac
 k 2) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Evening break
DTSTART:20241122T100000Z
DTEND:20241122T103000Z
DTSTAMP:20260419T170150Z
UID:session/HqMdJvgaf5jahtEg4ZsUfp@hasgeek.com
SEQUENCE:12
CREATED:20241110T030843Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T052732Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Evening break in Board Room\, 2nd floor (Security Engineering 
 Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Evening break
DTSTART:20241122T101000Z
DTEND:20241122T104000Z
DTSTAMP:20260419T170150Z
UID:session/C4C1VEmSVeBksZbboFz97y@hasgeek.com
SEQUENCE:3
CREATED:20241111T083228Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241114T151928Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Evening break in Gallery\, 2nd floor (Systems Engineering Trac
 k 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Open Source AI Community - kick-off meet-up
DTSTART:20241122T103000Z
DTEND:20241122T113000Z
DTSTAMP:20260419T170150Z
UID:session/AwAgYRixw1a8rp1YBD4x8M@hasgeek.com
SEQUENCE:13
CREATED:20241110T021809Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241120T091345Z
LOCATION:Board Room\, 2nd floor (Security Engineering Track) - Bangalore I
 nternational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Open Source AI Community - kick-off meet-up in Board Room\, 2n
 d floor (Security Engineering Track) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Re-imagining data reactivity: the story of DiceDB
DTSTART:20241122T104000Z
DTEND:20241122T111000Z
DTSTAMP:20260419T170150Z
UID:session/PisfuNEU3VWd26DeE5qUZr@hasgeek.com
SEQUENCE:14
CATEGORIES:40 min talk,Systems engineering
CREATED:20241109T163839Z
DESCRIPTION:The talk will deep dive into how we built DiceDB and made it a
  truly real-time reactive database by eliminating polling inefficiencies. 
 The talk will touch on the internal arch of Redis\, persistent connections
 \, and leveraging Pub/Sub patterns to enable instantaneous data flow at lo
 w latency and high throughput.\n\nThe talk will be focussed on the core de
 sign decisions of DiceDB that enabled us to build efficient query subscrip
 tions using `.WATCH` commands and the inner components like query fingerpr
 ints\, subscription maps\, and the database watcher. The difference in use
 r experience and ease of building will be demonstrated through a real-time
  leaderboard example.\n\nUltimately\, the audience will get a solid unders
 tanding of the foundational work that makes DiceDB real-time and reactive.
  Some key takeaways are\n\n- the internal architecture of Redis\n- challen
 ges with building 1:n fanout with poll-based system\n- how we re-architect
 ed it to make a push-based\n- key design principles behind `.WATCH` comman
 ds for query subscriptions and real-time data updates\n\nThis talk is idea
 l for backend engineers\, database enthusiasts\, system architects\, and t
 echnical leads working on real-time\, data-intensive applications. It will
  particularly resonate with engineers already familiar with Redis. It’s 
 also highly relevant for developers working on large-scale distributed sys
 tems or infrastructure teams looking to optimize performance and responsiv
 eness in high-concurrency environments.
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074228Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
URL:https://hasgeek.com/rootconf/2024/schedule/how-we-made-dicedb-a-truly-
 real-time-reactive-database-PisfuNEU3VWd26DeE5qUZr
BEGIN:VALARM
ACTION:display
DESCRIPTION:Re-imagining data reactivity: the story of DiceDB in Gallery\,
  2nd floor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Transition
DTSTART:20241122T111000Z
DTEND:20241122T112000Z
DTSTAMP:20260419T170150Z
UID:session/6aYEu1PrftbzC1PwKWu538@hasgeek.com
SEQUENCE:1
CREATED:20241111T084202Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241111T084227Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Transition in Gallery\, 2nd floor (Systems Engineering Track 1
 ) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
BEGIN:VEVENT
SUMMARY:Wrap-up session - Systems Engineering Track
DTSTART:20241122T112000Z
DTEND:20241122T114000Z
DTSTAMP:20260419T170150Z
UID:session/BdNudxmiY2t2tDPp4iiiJP@hasgeek.com
SEQUENCE:7
CREATED:20241111T084220Z
GEO:12.9317;77.6227
LAST-MODIFIED:20241225T074200Z
LOCATION:Gallery\, 2nd floor (Systems Engineering Track 1) - Bangalore Int
 ernational Centre\nBengaluru\nIN
ORGANIZER;CN=Rootconf:MAILTO:no-reply@hasgeek.com
BEGIN:VALARM
ACTION:display
DESCRIPTION:Wrap-up session - Systems Engineering Track in Gallery\, 2nd f
 loor (Systems Engineering Track 1) in 5 minutes
TRIGGER:-PT5M
END:VALARM
END:VEVENT
END:VCALENDAR
