# E2E Cloud Documentation

> E2E Cloud (E2E Networks) is an Indian cloud platform providing compute nodes, GPU instances, managed databases, Kubernetes, object storage, networking, security, and AI/ML infrastructure. The MyAccount portal at https://myaccount.e2enetworks.com is the primary surface for launching, connecting to, and operating resources; the documentation at https://docs.e2enetworks.com mirrors the portal product structure.

> To fetch the raw markdown of any page, append .md to the doc path, for example /docs/myaccount/node/getting-started/create-node.md

> Indexes are layered. This file lists product families. The MyAccount index at `/docs/myaccount/llms.txt` lists every product under the MyAccount console. Per-product indexes (for example `/docs/myaccount/node/llms.txt`, `/docs/myaccount/gpu/llms.txt`) list every page in that product with a one-line description.

## Product Families

- [MyAccount](https://docs.e2enetworks.com/docs/myaccount/llms.txt): The MyAccount cloud management console. Detailed index covers Compute (Nodes, GPU, Kubernetes, Auto Scaling, Functions, Container Registry), Storage (Block, Object, Parallel File Storage, DR, CDP Backups, Images), Database, Network (VPC, IPs, security groups, Load Balancers), Security and Compliance, Messaging and Observability, Account/Billing/Access, and Automation (CLI, Terraform).

## Detailed Product Indexes (TIR)

- [Slurm Cluster (TIR)](https://docs.e2enetworks.com/docs/tir/SlurmCluster/llms.txt): TIR's managed Slurm scheduler on Kubernetes (Slinky) for distributed GPU training - architecture (login node, controller, GPU workers, partitions), creating a cluster field by field, connecting over SSH and per-user Login User Management, running any Docker/OCI container image inside a Slurm job with Enroot and Pyxis (flags, image caching as squash files, multi-node NCCL training), extending slurm.conf and defining partitions and prolog/epilog scripts from the console, day-2 management across the ten details-page tabs plus every action and its blockers, a screen-by-screen UI guide, billing, and symptom-first troubleshooting.
- [Object Storage (TIR)](https://docs.e2enetworks.com/docs/tir/object-storage/llms.txt): S3-compatible EOS buckets, previously known as Datasets (EOS) - creating a bucket (global name namespace, alias vs bucket name, encryption, create-time object lock), uploading from the portal/mc/s3cmd, object versioning and per-version governance lock and legal hold, lifecycle expiry rules, presigned URLs, cross-region replication, peak-usage slab pricing, and symptom-first troubleshooting.
- [Block Storage (TIR)](https://docs.e2enetworks.com/docs/tir/block-storage/llms.txt): fixed-size PVC disks, previously known as Datasets (Disk) - creating one (10 GB minimum, 100 GB default, project storage limit), mounting read-write at /datasets/<name> on one resource at a time, growing a disk and requesting a higher limit, allocated-size pricing against Object Storage's peak-usage model, and symptom-first troubleshooting.

## Detailed Product Indexes (MyAccount)

- [Nodes](https://docs.e2enetworks.com/docs/myaccount/node/llms.txt): Compute nodes - planning and launching CPU/Marketplace/Private Cluster nodes, connecting via SSH/RDP, day-2 operations (actions, security, network, action availability), capabilities (E1, images, snapshots, encryption, monitoring), and symptom-first troubleshooting.
- [GPU](https://docs.e2enetworks.com/docs/myaccount/gpu/llms.txt): GPU-backed compute nodes - choosing an NVIDIA card (H100, A100, L40S, A40, A30, L4, V100, T4), launching a GPU node (the create flow inverts to GPU-first then OS/plan), connecting and verifying with `nvidia-smi`, GPU-specific management differences (per-card monitoring, Save Image of driver+CUDA, plan-upgrade restrictions, no snapshots), hands-on guides (Docker with `--gpus all`, baking and reusing a GPU image, serving vLLM/TGI inference), and GPU-specific troubleshooting.
- [Load Balancer](https://docs.e2enetworks.com/docs/myaccount/load-balancer/llms.txt): Managed load balancers - choosing between an Application Load Balancer (ALB, Layer 7 HTTP/HTTPS with ACL routing, health checks, and auto-scaling backends) and a Network Load Balancer (NLB, Layer 4 TCP), creating either type, day-2 management across the Info, Backend Mapping, ACL, Network, Monitoring/Alerts/Stats, Security Group, and logs/timeline tabs plus lifecycle actions and committed billing, serving HTTPS with imported SSL certificates, encryption at rest, and symptom-first troubleshooting.