Release v1.6.0

Fixes

• Correct a statement about multiple exports in the "nfs create" help text. (ba82ed5)

Release v1.6.0-rc.1

Changes

• Add a sensible User-Agent when sending http requests via the client. (fc34f08)
• NFS: Let LINSTOR set the file system owner of the export's root directory. This allows XFS file systems to be used with NFS exports. (01161bf)
• NFS: Add a --filesystem option when creating an NFS export. (86b70a5)

Fixes

• Fix some Go API client functions so that they actually return data. (34f68fc)

Release v1.5.0

Features

• Added hints to some health check outputs to help users resolve potential issues. (09e0860)
• New health check for NVMe-oF resource agents. (d3441f5)
• New health check for SCST kernel modules and user space utilities. (138c1f6)
• The drbd-reactor health check now also checks that if the drbd-reactor-reload.{path,service} files are up-to-date. (2c40ff0)
• When creating a resource, linstor-gateway now waits a few seconds to ensure that the resource is actually running. This prevents situations where the resource fails to start, but linstor-gateway still reports it as "running". (2c40ff0)
• Various improvements to the end-to-end tests.

Release v1.5.0-rc.1

Features

• Added hints to some health check outputs to help users resolve potential issues. (09e0860)
• New health check for NVMe-oF resource agents. (d3441f5)
• New health check for SCST kernel modules and user space utilities. (138c1f6)
• The drbd-reactor health check now also checks that if the drbd-reactor-reload.{path,service} files are up-to-date. (2c40ff0)
• When creating a resource, linstor-gateway now waits a few seconds to ensure that the resource is actually running. This prevents situations where the resource fails to start, but linstor-gateway still reports it as "running". (2c40ff0)
• Various improvements to the end-to-end tests.

Release v1.4.0

• No changes over rc1

v1.4.0-rc.1

Features

• The nfs create command now supports creating multiple NFS exports in the same cluster. (6244def)
• The delete command now prompts for confirmation before deleting a resource. (7c6d783)
• drbd-reactor configuration files generated by LINSTOR Gateway are now versioned and can be migrated to new versions automatically. (8723022, f72c2c3)

Fixes

• The nfs upgrade command now actually respects the --dry-run flag. (e760b41)
• drbd-reactor configuration files no longer include the deprecated id field. (878d53f)
• In the NFS configuration file, the IPaddr2 and nfsserver resource agent entries were switched. These are now in the correct order. (2337e32)

v1.3.0

Features

• The status command will now show what node a resource is running on. (106dfc8)
• The check-health command now differentiates between "agent", "server", and "client" mode, checking the requirements for the respective mode of operation. (e0b2507)
• All create commands now use -r as the shorthand flag for resource groups. Previously, iSCSI used -g instead -- this shorthand is now deprecated. (5b0eb9a)
• When creating an iSCSI target, the --implementation flag can now be used to specify what iSCSI target implementation should be used by the underlying resource agent. (e6bcc14)
• When a step in the creation of a resource fails, the resources created in previous steps will now be cleaned up. This avoids "resource already exists" errors. (434d2d1)
• When installing LINSTOR Gateway via rpm, the drbd-reactor auto-reload files will now be copied to the appropriate place automatically. (736cd2a)

Fixes

• The check-health command now finds the correct path to the drbd-reactor documentation directory more reliably. (bc4747a)

Release v1.2.0

Features

• When creating a resource (iSCSI, NVMe-oF, or NFS), LINSTOR Gateway will now do a basic check for IP collisions. If the provided service IP is already in use by another resource, an error will be thrown. (48ac850)

Fixes

• The CLI will no longer print the usage information whenever an error happens. (4e2cc19)
• When listing resources, degraded resources could previously be displayed twice. This was changed so that every resource only gets reported once. (a4b56ce)
• Remove a check that verifies a resource is stopped before adding a volume. Adding volumes works just as well on running resources. (8426c36)

Release v1.1.1

Release v1.1.1

Release v1.1.0

Release v1.1.0