Alex Glikson

Info

• e-mail: aglikson@andrew.cmu.edu
• Role: Staff Engineer at Computer Science Department, Carnegie Mellon University (since S18)
• Research Interests: Cloud Infrastructure, Cloud Platforms, Serverless, Resource Management, Internet of Things, Edge Computing
• Office: GHC 7008
• My profile at the Parallel Data Lab (PDL)
• Social: LinkedIn, Twitter, Medium, SlideShare, Google Scholar

Teaching

• 15-719 - Advanced Cloud Computing (S18, S19; PhD level)
• 15-689 - Independent Study in Computer Sciences (F18; Master level)
• MCDS Capstone Project (F18; Master level)
• 15-890 - M.Sc. Graduate Project (S19; Master level)

Projects and Advisory

• Advanced Cloud Computing (15-719)
  • Advanced auto-scaling of Deep Learning model inference workloads on AWS
    • Dev TAs/RAs: Mingqiang Liang (M18), Hanqing Chen (M18, F18), Yupeng Zhang (F18)
  • Large-scaling simulation of job scheduling with Kubernetes
    • Dev TAs/RAs: Mingqiang Liang (M18), Hanqing Chen (M18)
• Cloud Platforms Research
  • Efficient on-demand 0-1 scaling of stateful interactive cloud-native applications on Kubernetes
    • Advisee: Shichao Nie (M.Sc. Graduate Project - S19)
    • Use-cases: Interactive Shell, Remote build/execution with IntelliJ IDE, Jupyter Notebook (bash, Python)
    • Talk at KubeCon Europe 2019
  • Rapid elasticity of Deep Learning model inference workloads by reducing 'cold' starts in Knative serverless platform (report).
    • Advisee: Ping-Min Lin (M.S. Independent Study - F18)
  • Enabling GPUs for rapidly elastic Deep Learning model inference workloads running on Knative
    • Advisee: Shengjie Luo (MCDS Capstone - F18)
  • Adaptive vertical elasticity of event-driven serverless workloads having resource demand correlated with event metadata
    • Advisee: Ke Chang (MCDS Capstone - F18)
• Cloud platforms enabling Technology-Enhanced Learning (TEL)
  • TheProject.Zone
    • Autograding of cloud computing projects
    • Collaborator: prof Majd Sakr
  • Diderot
    • Hosting of in-house online interactive learning environment on AWS (Elastic Beanstalk (Django), EC2, EBS, ELB, Auto Scaling, RDS, CloudFormation, Route53, Certificate Manager, S3, SES, SNS, Batch, ECS, ECR, Lambda, Kinesis Firehose, Glue, Athena, IAM, CloudWatch)
    • Collaborator: prof Umut Acar
  • ExamSystem
    • Multi-tenant, self-service hosting of in-house ExamSystem web application (Docker, Docker-Compose, scripts)
    • Collaborator: prof Seth Goldstein

Past Projects

• Migrating 'legacy' interactive applications to AWS Lambda (blog)
  • Collaborator: prof Guy Blelloch

Students

• Shichao Nie (M.Sc. Graduate Project; S19)
• Hanqing Chen (15-719 projects; M18, F18)
• Ping-Min Lin (IS, F18)
• Yupeng Zhang (15-719 project, F18)
• Shengjie Luo (Capstone, F18)
• Ke Chang (Capstone, F18)
• Mingqiang Liang (15-719 projects; M18)