Security and privacy at Modal

Добавлено

• Добавлено: Employee laptops are enrolled in mobile device management (MDM) with full disk encryption enforced.
• Добавлено: Network and infrastructure security (InfraSec) InfraSec covers how we secure the infrastructure that runs your workloads.
• Добавлено: We continuously monitor platform logs and metrics through third-party observability providers.
• Добавлено: Each container runs in its own sandbox, isolated from the host, using primitives like gVisor or microVMs.

• Добавлено: We remediate vulnerabilities in Modal’s systems within the timeframes below, measured from when a fix becomes available. Severity is based on the CVSS rating of the vulnerability and our assessment of its impact on Modal.
• Добавлено: We welcome responsible disclosure from the security community and run a private bug bounty program through HackerOne. To participate, email security@modal.com with your HackerOne username and we will send an invite. When performing security research, you must use a Modal Workspace whose name ends in -H1-, where is your HackerOne username.
• Добавлено: This section covers how long each type of data is kept, which products retain no data at all, and where your data is stored.
• Добавлено: Retention varies by product; the table below shows how long each type of data is retained. All stored data is encrypted at rest.
• Добавлено: Note
• Добавлено: While we store app logs and metadata as part of operating the platform, we access them only with your permission, to help troubleshoot an issue.
• Добавлено: Zero data retention
• Добавлено: Dedicated and Shared inference endpoints have zero data retention. Request and response payloads are never written to disk and pass through Modal only as in-flight network traffic.
• Добавлено: Data residency
• Добавлено: See our data residency guide for where each type of data is stored and the controls available for residency requirements.
• Добавлено: Shared responsibility model
• Добавлено: Modal prioritizes the integrity, security, and availability of customer data. Under our shared responsibility model, you also have responsibilities in the areas below.
• Добавлено: AreaModalCustomer Access and secretsProvides access controls, including SSO, SCIM, API tokens, and RBAC, and Secrets management for storing credentials.Manage the identities in your Workspace, assign roles, and rotate and remove API tokens. Own the contents and rotation of your Secrets.Encryption and network securityEncrypts data in transit with TLS 1.3 and at rest.Decide which endpoints you expose and how they are authenticated. Apply any additional encryption your data requires, such as encrypting sensitive fields before they reach Modal.VulnerabilitiesPatches the platform and runtimes within our severity timeframes and audits our dependencies for known vulnerabilities.Patch your Images, dependencies, and code.Untrusted codeProvides isolation primitives for running untrusted code through Restricted Functions and Sandboxes.Run untrusted code, such as LLM-generated or end-user-submitted code, using those primitives and with guardrails such as egress restrictions, resource limits, and timeouts. Keep Secrets and credentials out of untrusted workloads.Data lifecycleEnsures the durability of managed storage and applies our retention and deletion policies.Maintain backups of the data you store in Modal and routinely verify their integrity.OperationsOperates the platform for high availability, monitors it, and responds to platform incidents.Monitor your applications and design for failover.ComplianceMakes our audit reports and control documentation available on our Trust Center.Determine which laws and regulations apply to your organization and your data, and configure and use Modal to meet them.
• Добавлено: Security features
• Добавлено: We provide security features across our products to help you secure your workloads, such as single sign-on, Role-Based Access Control (RBAC), Sandbox network access controls, audit logs, and customer-supplied encryption keys.
• Добавлено: ProductFeatures WorkspaceBlock unauthenticated endpoints, Role-Based Access Control (RBAC), Proxies (static egress IPs)FunctionsRestricted FunctionsSandboxesOutbound access control, Inbound access controlIdentityOkta SSO, Microsoft Entra SSO, Custom SAML SSO, SCIM integration, OIDC integrationObservabilityAudit logs, Datadog integration, OpenTelemetry integrationEncryptionCustomer-supplied encryption keys

Изменено

• Изменено: Было Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources Sidecars BetaSecret injection BetaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Training Getting started Tinker-compatible API Multi -node training Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node Clusters Deployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Sticky Sessions Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups BetaRole -Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics Стало Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources Sidecars BetaSecret injection BetaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint candidates Endpoint integrations Training Quickstart Tinker-compatible API BetaMulti -node training Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node Clusters Deployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Sticky Sessions Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration Slack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups Role -Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics
• Изменено: Было We conduct annual business continuity and security incident exercises. Стало We run business continuity and security incident exercises every year .

Удалено

• Удалено: The document outlines Modal’s security and privacy commitments.
• Удалено: Application security (AppSec)
• Удалено: AppSec is the practice of building software that is secure by design, secured
• Удалено: during development, secured with testing and review, and deployed securely.

• Удалено: We build our software using memory-safe programming languages, including Rust
• Удалено: (for our worker runtime and storage infrastructure) and Python (for our API
• Удалено: servers and Modal client).
• Удалено: Software dependencies are audited by Github’s Dependabot.
• Удалено: We make decisions that minimize our attack surface. Most interactions with
• Удалено: Modal are well-described in a gRPC API, and occur through modal, our open-source command-line tool
• Удалено: and Python client library.
• Удалено: We have automated synthetic monitoring test applications that continuously
• Удалено: check for network and application isolation within our runtime.
• Удалено: We use HTTPS for secure connections. Modal forces HTTPS for all services using
• Удалено: TLS (SSL), including our public website and the Dashboard to ensure secure
• Удалено: connections. Modal’s client library connects
• Удалено: to Modal’s servers over TLS and verify TLS certificates on each connection.
• Удалено: All user data is encrypted in transit and at rest.
• Удалено: All public Modal APIs use TLS 1.3, the latest and
• Удалено: safest version of the TLS protocol.
• Удалено: Internal code reviews are performed using a modern, PR-based development
• Удалено: workflow (Github), and engage external penetration testing firms to assess our
• Удалено: software security.
• Удалено: Corporate security (CorpSec)
• Удалено: Employee laptops are protected by full disk encryption using FileVault2, and
• Удалено: managed by Secureframe MDM.
• Удалено: Network and infrastructure security (InfraSec)
• Удалено: InfraSec is the practice of ensuring a hardened, minimal attack surface for
• Удалено: components we deploy on our network.
• Удалено: Modal uses logging and metrics observability providers, including Datadog and
• Удалено: Sentry.io.
• Удалено: Compute jobs at Modal are containerized and virtualized using gVisor, the sandboxing technology
• Удалено: developed at Google and used in their Google Cloud Run and *Google
• Удалено: Kubernetes Engine* cloud services. Sandboxes may also run on Modal’s
• Удалено: secure VM runtime.
• Удалено: Security vulnerabilities directly affecting Modal’s systems and services will be
• Удалено: patched or otherwise remediated within a timeframe appropriate for the severity
• Удалено: of the vulnerability, subject to the public availability of a patch or other
• Удалено: remediation mechanisms.
• Удалено: If there is a CVSS severity rating accompanying a vulnerability disclosure, we
• Удалено: rely on that as a starting point, but may upgrade or downgrade the severity
• Удалено: using our best judgement.
• Удалено: Shared responsibility model
• Удалено: Modal prioritizes the integrity, security, and availability of customer data. Under our shared responsibility model, customers also have certain responsibilities regarding data backup, recovery, and availability.
• Удалено: Data backup: Customers are responsible for maintaining backups of their data. Performing daily backups is recommended. Customers must routinely verify the integrity of their backups.
• Удалено: Data recovery: Customers should maintain a comprehensive data recovery plan that includes detailed procedures for data restoration in the event of data loss, corruption, or system failure. Customers must routinely test their recovery process.
• Удалено: Availability: While Modal is committed to high service availability, customers must implement contingency measures to maintain business continuity during service interruptions. Customers are also responsible for the reliability of their own IT infrastructure.
• Удалено: Security measures: Customers must implement appropriate security measures, such as encryption and access controls, to protect their data throughout the backup, storage, and recovery processes. These processes must comply with all relevant laws and regulations.
• Удалено: SOC 2
• Удалено: We have successfully completed a System and Organization Controls (SOC) 2 Type 2
• Удалено: audit. Go to our Security Portal to request access to the report.
• Удалено: HIPAA
• Удалено: HIPAA, which stands for the Health Insurance Portability and Accountability Act, establishes a set of standards that protect health information, including individuals’ medical records and other individually identifiable health information. HIPAA guidelines apply to both covered entities and business associates—of which Modal is the latter if you are processing PHI on Modal.
• Удалено: Modal’s services can be used in a HIPAA compliant manner. It is important to note that unlike other security standards, there is no officially recognized certification process for HIPAA compliance. Instead, we demonstrate our compliance with regulations such as HIPAA via the practices outlined in this doc, our technical and operational security measures, and through official audits for standards compliance such as SOC 2 certification.
• Удалено: To use Modal services for HIPAA-compliant workloads, a Business Associate Agreement (BAA) should be established with us prior to submission of any PHI. This is available on our Enterprise plan. Contact us at security@modal.com to get started. At the moment, Volumes v1, Images (excluding Filesystem and Directory Snapshots), Memory Snapshots, and user code are out of scope of the commitments within our BAA, so PHI should not be used in those areas of the product.
• Удалено: Volumes v2 are HIPAA compliant.
• Удалено: PCI
• Удалено: Payment Card Industry Data Security Standard (PCI) is a standard that defines
• Удалено: the security and privacy requirements for payment card processing.
• Удалено: Modal uses Stripe to securely process transactions and
• Удалено: trusts their commitment to best-in-class security. We do not store personal
• Удалено: credit card information for any of our customers. Stripe is certified as “PCI
• Удалено: Service Provider Level 1”, which is the highest level of certification in the
• Удалено: payments industry.
• Удалено: Keeping user data secure is a top priority at Modal. We welcome contributions
• Удалено: from the security community to identify vulnerabilities in our product and
• Удалено: disclose them to us in a responsible manner. We currently run a private bug
• Удалено: bounty program through HackerOne. If you have found a vulnerability and
• Удалено: wish to participate, please send an email to security@modal.com with your
• Удалено: HackerOne username or email and we will invite you to the program.
• Удалено: Modal will never access or use:
• Удалено: your source code.
• Удалено: the inputs (function arguments) or outputs (function return values) to your Modal Functions.
• Удалено: any data you store in Modal, such as in Images or Volumes.
• Удалено: App logs and metadata are stored on Modal. Modal will not access this data
• Удалено: unless permission is granted by the user to help with troubleshooting.
• Удалено: Different Modal products have different retention policies for the data they
• Удалено: handle. The table below summarizes how long each type of data is retained.
• Удалено: Function inputs and outputs
• Удалено: Function inputs and outputs are stored encrypted at rest. Small payloads
• Удалено: (≤ 2 MiB) are stored inline in our metadata store; larger payloads are stored
• Удалено: in object storage. Both are deleted within a maximum TTL of 7 days.
• Удалено: Modal Inference endpoints
• Удалено: Modal Inference endpoints are zero data retention: request and response payloads
• Удалено: are never written to disk and pass through Modal’s infrastructure only as
• Удалено: in-flight network traffic. Inference endpoints terminate TLS at Modal’s edge
• Удалено: proxy and forward requests directly to your containers over an internal tunnel.
• Удалено: Questions?
• Удалено: Email us!
• Удалено: Security and privacy at ModalApplication security (AppSec)Corporate security (CorpSec)Network and infrastructure security (InfraSec)Vulnerability remediationSeverity timeframesShared responsibility modelSOC 2HIPAAPCIBug bounty programData privacyData retentionFunction inputs and outputsModal Inference endpointsQuestions?

Изменено

• Изменено: Было Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources Sidecars BetaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Training Getting started Tinker-compatible API Multi-node training Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node Clusters Deployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Sticky Sessions Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups BetaRole-Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics Стало Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources Sidecars BetaSecret injection BetaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Training Getting started Tinker-compatible API Multi-node training Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node Clusters Deployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Sticky Sessions Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups BetaRole-Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics

Добавлено

• Добавлено: Kubernetes Engine* cloud services. Sandboxes may also run on Modal’s
• Добавлено: secure VM runtime.

Изменено

• Изменено: Было Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources VM Sandboxes BetaSidecars AlphaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node clusters BetaDeployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups BetaRole-Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics Стало Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources Sidecars BetaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Training Getting started Tinker-compatible API Multi-node training Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node Clusters Deployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Sticky Sessions Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups BetaRole-Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics

Удалено

• Удалено: Kubernetes Engine* cloud services.

Изменено

• Изменено: Было Introduction Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources VM Sandboxes BetaSidecars AlphaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node clusters BetaDeployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Modal user account setup Service users User groups BetaRole-Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics Стало Guide Introduction to Modal Getting started Functions Functions Invocation methods Function lookups Container lifecycle management Parametrized Functions Dynamic configuration Sandboxes Sandboxes Running commands Networking and security File access Snapshots Pricing and resources VM Sandboxes BetaSidecars AlphaEndpoints Endpoints Shared Endpoints Dedicated Endpoints Endpoint integrations Images Defining Images Using existing container images Named images Fast pull from registry GPUs and other resources GPU acceleration Using CUDA on Modal Configuring CPU, memory, and disk Scaling out Scaling out Input concurrency Batch processing Job queues Dynamic batching Multi-node clusters BetaDeployment Apps, Functions, and entrypoints Managing deployments Continuous deployment Scheduled Functions HTTP Applications Servers Web Functions Streaming endpoints Web Function URLs Request timeouts Proxy Tokens Networking Tunnels Proxies BetaCluster networking Data sharing and storage Passing local data Volumes Storing model weights Cloud bucket mounts Dicts Queues Dataset ingestion Secrets and environment variables Secrets Environment variables Performance Cold start performance Memory Snapshots High-performance LLM inference Reliability and robustness Failures and retries Preemption Timeouts GPU health Troubleshooting Security and privacy Security and privacy Data residency Customer-supplied encryption keys AlphaAudit logs Integrations Using OIDC to authenticate with external services Connecting Modal to your Datadog account Connecting Modal to your OpenTelemetry provider Okta SSO Microsoft Entra SSO Custom SAML SSO SCIM Integration BetaSlack notifications BetaWorkspace & account settings Workspaces Environments Service users User groups BetaRole-Based Access Control (RBAC) Billing Network egress billing Budgets Other topics Feature maturity Notebooks Restricted Functions JavaScript/Go SDKs BetaModal 1.0 migration guide File and project structure Developing and debugging Jupyter notebooks Asynchronous API usage Global variables Region selection GPU Metrics