v1.1.1

NVIDIA Fleet Intelligence Agent v1.1.1

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

What's Changed

• fix: release 1.1 security fixes by @jingxiang-z in #164
• fix: more security fix for release 1.1 by @jingxiang-z in #167
• docs: add sudo to unix socket curl examples (#174) by @rsampaio in #175
• security: disable HTTP redirect following on attestation, enrollment, and CLI clients [1/5] (#169) by @jingxiang-z in #177

Full Changelog: v1.1.0...v1.1.1

v1.1.1-rc.3

NVIDIA Fleet Intelligence Agent v1.1.1-rc.3

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

Other Changes

• bf444fd: security: disable HTTP redirect following on attestation, enrollment, and CLI clients [1/5] (#169) (#177) (@jingxiang-z)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v1.2.0-rc.4

NVIDIA Fleet Intelligence Agent v1.2.0-rc.4

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

What's Changed

• feat: add dcgm version to machine info by @jingxiang-z in #143
• fix: improve goreleaser tag selection and changelog grouping by @jingxiang-z in #144
• fix: secure fleetint state permissions by @jingxiang-z in #147
• chore: remove unused gpu-count flag and improve configuration docs by @jingxiang-z in #146
• fix: dcgm incident integer types by @jingxiang-z in #148
• fix: restore localhost-only server defaults by @jingxiang-z in #151
• fix: validate derived backend endpoints by @jingxiang-z in #152
• fix: bump sdk grpc-go by @jingxiang-z in #155
• fix: harden startup and export under slow DCGM and machine info by @jingxiang-z in #157
• fix: detect gpu hardware before driver checks by @jingxiang-z in #158
• chore: add SECURITY.md and bump Go to 1.26.2 by @jingxiang-z in #159
• chore: prune unused vendored sdk packages by @jingxiang-z in #161
• security: remove dead code fix cmd exec by @rsampaio in #162
• feat: Add event_id(UUID) to events info in agent message to backend by @ambermingxin in #160
• security: harden agent against local and network attack vectors by @rsampaio in #163
• fix: harden exporter JWT refresh handling by @jingxiang-z in #165
• fix: enforce token file size limit before read by @rsampaio in #166
• fix: harden exporter redirects for custom HTTP clients by @jingxiang-z in #168
• docs: add sudo to unix socket curl examples by @rsampaio in #174
• security: disable HTTP redirect following on attestation, enrollment, and CLI clients [1/5] by @rsampaio in #169
• security: use absolute paths for exec to prevent PATH poisoning [2/5] by @rsampaio in #170
• security: sanitize error responses and clamp query time ranges [3/5] by @rsampaio in #171
• security: add security response headers to HTTP API [4/5] by @rsampaio in #172

Full Changelog: v1.1.0...v1.2.0-rc.4

v1.1.1-rc.2

NVIDIA Fleet Intelligence Agent v1.1.1-rc.2

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

Documentation Updates

• 78609f9: docs: add sudo to unix socket curl examples (#174) (#175) (@rsampaio)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v1.1.1-rc.1

NVIDIA Fleet Intelligence Agent v1.1.1-rc.1

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

Bug Fixes

• 046e78e: fix: enforce token file size limit before read (#166) (@rsampaio)
• 07abe38: fix: harden exporter JWT refresh handling (#165) (@jingxiang-z)
• 4cc9037: fix: harden exporter redirects for custom HTTP clients (@jingxiang-z)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v1.1.1-rc.0

NVIDIA Fleet Intelligence Agent v1.1.1-rc.0

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

Bug Fixes

• b5e9e18: fix: restore localhost-only server defaults (#151) (@jingxiang-z)
• 9cb8262: fix: secure fleetint state permissions (#147) (@jingxiang-z)
• 65819c5: fix: validate derived backend endpoints (#152) (@jingxiang-z)

Other Changes

• 16e6731: security: harden agent against local and network attack vectors (#163) (@rsampaio)
• d2c2b8b: security: remove dead code fix cmd exec (#162) (@rsampaio)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v0.0.1-poc.otelgateway

NVIDIA Fleet Intelligence Agent v0.0.1-poc.otelgateway

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

Other Changes

• da8306d: [GPUHEALTH-1520] feat: add gpu index to machine info (#132) (@ambermingxin)
• 313b24e: [GPUHEALTH-1556] Change to Prod URL and remove deployEnv (@mukilsh)
• 2767124: [GPUHEALTH-1563] docs: fix the xid component name (@jingxiang-z)
• aed2cb6: add header to missing file (@rsampaio)
• d608fa2: feat: add OTel gateway collector with SAK auth extension (@rsampaio)
• 4ffbbeb: feat: add enrollment precheck command (#134) (@jingxiang-z)
• cdbfc36: feat: add typed health incidents (#137) (@jingxiang-z)
• fd2ecda: fix: export NVLink BER float metrics (#140) (@jingxiang-z)
• 6806019: fix: remove retired page metrics (@jingxiang-z)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v1.2.0-rc.3

NVIDIA Fleet Intelligence Agent v1.2.0-rc.3

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

Bug Fixes

• 9c1c3a0: fix: bound dcgm startup during server init (@jingxiang-z)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v1.2.0-rc.2

NVIDIA Fleet Intelligence Agent v1.2.0-rc.2

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

New Features

• 7c5fc96: feat: add dcgm version to machine info (#143) (@jingxiang-z)

Bug Fixes

• 465d156: fix: avoid repeated initial machine info wait (@jingxiang-z)
• f86cfff: fix: bump sdk grpc-go (#155) (@jingxiang-z)
• 7d29806: fix: cancel initial machine info wait with context (@jingxiang-z)
• 41e9c3b: fix: dcgm incident integer types (#148) (@jingxiang-z)
• 1546195: fix: harden export under slow machine info collection (@jingxiang-z)
• 2fa2d5d: fix: improve goreleaser tag selection and changelog grouping (#144) (@jingxiang-z)
• cd41477: fix: restore localhost-only server defaults (#151) (@jingxiang-z)
• d08aead: fix: secure fleetint state permissions (#147) (@jingxiang-z)
• 752b61f: fix: validate derived backend endpoints (#152) (@jingxiang-z)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues

v1.2.0-rc.1

NVIDIA Fleet Intelligence Agent v1.2.0-rc.1

This release includes a host agent for GPU telemetry collection and attestation.

Available Packages

• DEB packages: amd64 and arm64 for Debian/Ubuntu systems
• RPM packages: x86_64 and aarch64 for RHEL/CentOS/Fedora systems
• Binary archives: Standalone binaries for both architectures

Installation

Download the appropriate package for your system from the assets below. For installation instructions, see the project README.

Changelog

New Features

• 7c5fc96: feat: add dcgm version to machine info (#143) (@jingxiang-z)

Bug Fixes

• 41e9c3b: fix: dcgm incident integer types (#148) (@jingxiang-z)
• 1372c70: fix: decouple export upload and machine info collection (@jingxiang-z)
• 2fa2d5d: fix: improve goreleaser tag selection and changelog grouping (#144) (@jingxiang-z)
• af43c37: fix: isolate exporter timeout phases (@jingxiang-z)
• cd41477: fix: restore localhost-only server defaults (#151) (@jingxiang-z)
• d08aead: fix: secure fleetint state permissions (#147) (@jingxiang-z)
• 752b61f: fix: validate derived backend endpoints (#152) (@jingxiang-z)

Verification

All release artifacts are checksummed. Verify downloads using:

sha256sum -c checksums.txt

Support

For issues and questions, please visit: https://github.com/NVIDIA/fleet-intelligence-agent/issues