Releases: LINBIT/linstor-server
v1.2.1
linstor-server 1.2.1
- Allow configuration of etcd with tls
- Use more environment variables to configure Controller or Satellite:
LS_REST_BIND_ADDRESS
LS_REST_BIND_ADDRESS_SECURE
LS_CONFIG_DIRECTORY
LS_LOG_DIRECTORY
LS_LOG_LEVEL
LS_KEEP_RES - DRBD resource files will now be backuped
- Delete old snapshot-restore properties that were never deleted
- Always skip initial sync on ZFS
- Fix query-max-volume-size REST call, expected incorrect POST data
- REST-API at version 1.0.10
v1.2.0
linstor-server 1.2.0
- Add auto-tiebreaker feature
- Fix possible deadlock on closeConnection event
- Better volume error reporting
- Fix a race condition on deleting peers
- Fixes to snapshot restoring
- Most client list filter now on the Controller
- NVME allow multiple NVMe initiators
- REST-API at version 1.0.10
v1.1.2
v1.1.1
linstor-server 1.1.1
- Node reconnect now also reconnect failed nodes
- Fixed dmstats command calling
- Reduced classpath for better process view
- Added exception handling for device handler (killing satellite in some cases)
- Correctly cleanup(remove from database) on retry-task operations
- NVME fixed ArrayIndexOutOfBoundsException if there are external NVME resources
- Added user name to log entries
- Fixed incomplete migration for external names
v1.1.0
ETCD support
There are some limitations regarding the ETCD support:
- There is no migration from SQL -> ETCD
- Linstor requires the ETCD server to increase its limit for requests per transactions [1]
Our tests were running well with 1024, but some linstor-actions are depending on how many objects you already have (i.e. auto-place)
you might want to increase that value if necessary.
example: etcd --max-txn-ops 1024
To use ETCD simply change your linstor.toml to contain
[db]
connection_url = "etcd://127.0.0.1:2379"
Some more release details:
linstor-server 1.1.0
- Add truststore to config and HTTPS rest-api
- StorPoolData: correclty set snaphot supported field
- IntAuthResponse: Fixed version mismatch impl error
- satellite: parse satellite settings from toml file
- Satellite nodes don't need anymore to have matching hostnames for DRBD to work
v1.0.1
v1.0.0
- Resource groups
A resource group is a parent object of a resource definition and all property changes
made on a resource group will be inherited by its resource definition children.
The resource group also stores settings for auto-place rules and can spawn a resource definition depending on the stored rules.
Here is a example:
linstor resource-group create my_group --place-count 3
linstor volume-group create my_group
linstor resource-group spawn-resources my_group myresource 2G
Some more release details:
linstor-server 1.0.0
- Added resource groups/volume groups
- REST-API version 1.0.8
- Fixed incorrect HTTP initialization if rest bind address was given by cli
- Fixed rollback in props container
- Fixed sysfs throttling with nvme diskless
- Allow log level to be set in linstor.toml or cli
- External command timeout is now configurable
v0.9.13
Release highlights:
- Search domain for nodes
If you have registered your linstor satellites with the FQDN you can now set the
new controller property SearchDomain with your common domain part and than
you only have to use your subdomain part for commands.
- File based storage pools (loop devices)
We have a new storage pool types file and file-thin that uses files as loop devices (via losetup), that support all of our Linstor features.
- HTTPS for REST-API
The REST-API can now be access through HTTPS if configured and all requests to HTTP will than return a redirect to the HTTPS address.
To setup HTTPS you have to create a java keystore file[1] with a valid certificate and
configure the [https] section in /etc/linstor/linstor.toml file.
[https]
keystore = "/etc/linstor/keystore_server.jks"
keystore_password = "linstor"
- Basic LDAP support
Linstor has now support for basic LDAP authentication, this means you can configure a LDAP server and a search_filter and allow only members of a certain group access to Linstor.
A requirement for LDAP is authentication support is that you have configured Linstor with HTTPS.
Here is a sample config entry in our new /etc/linstor/linstor.toml config file:
[ldap]
enabled = true
uri = "ldaps://ldap.example.com"
dn = "uid={user},ou=users,o=ha,dc=example"
search_base = "dc=example"
search_filter = "(&(uid={user})(memberof=cn=linstor,ou=services,o=ha,dc=example))"
The {user} template variable will be replaced with the login user.
- External meta-data support for DRBD
Before creating the resource just set the new StorPoolNameDrbdMeta property on the resource-definition or volume-definition for the storage pool you want to use for external meta-data.
- New TOML based configuration file
The current /etc/linstor/database.cfg should automatically migrated to the new /etc/linstor/linstor.toml file.
linstor-server 0.9.13
- Maria db with large index is now supported
- Basic LDAP support
- File based storage pools (FILE provider)
- External meta-data support for DRBD
- HTTPS support for REST-API
- REST-API version 1.0.7 (2 new fields)
- Node name resolving based on search domain property
- Fix always add resources to the retry task if they failed
- Improved error handling on REST-API calls
- Fixed LVM-thick allocation sizes
- Better show active net interface on nodes
v0.9.12
linstor-server 0.9.12
- Protobuf-API was removed, client 0.9.3 will NOT work anymore
- Added: modify volume rest uri, REST-API version 1.0.6
- Added: /health HTTP GET entry point usable for kubernetes liveness checks
- Allow throttling of IO fia cgroups blkio
- Enabled NVME on top of LUKS layer
- Fixed: Access to deleted data errors on satellite on resource delete
- Fixed: Hostname check on satellites using DRBD
- Fixed: list error reports if error report is from the future (time mismatches)
- Fixed: Allow PrefNic property on nodes
- Fixed: Incorrect allocation size on lvmthin
v0.9.11
linstor-server 0.9.11
- New: Archive logs that are older than 2 months
- Improved controller and satellite startup by ~800ms
- Better error reporting on REST-API errors
- Fixed: Unable to create new resource if there are resource-definitions without resources (introduced with 0.9.0)
- Fixed: Full-sync will not fail on broken storage pools
- Fixed: Storage pool error will be reported in new
reportsfield
on storage-pool objects (REST-API 1.0.5) - Fixed: Null pointer exception on snapshots with storage-only snapshots