Skip to content

Latest commit

 

History

History
356 lines (275 loc) · 15.6 KB

CHANGELOG.md

File metadata and controls

356 lines (275 loc) · 15.6 KB
Raw

Splunk Platform Automator changes by release

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog.

2.2.4 - 2024-12-27

Fixed

  • Disable COM1 when using Virtualbox on WSL with vagrant

Added

  • Added splunk_architecture to support different OS architectures like amd64, x86_64, arm64, etc.
  • Added splunk_fips to support FIPS mode in splunk

2.2.3 - 2024-12-01

Fixed

  • Forced splunk restart handler to the inventory_hostname to prevent mistaken delegations

Added

  • Added org_indexer_volume_indexes and org_all_indexes to heavy forwarders for app wizards index selection
  • Added requirements.txt to the project for easy module installation
  • Added commented out profiling for easy timestamp activation in the ansible output
  • Added support for vagrant-gecko-aws plugin

2.2.2 - 2024-03-05

Fixed

  • Fix the usage of the manager-apps dir on the cm from 9.0.0 on

2.2.1 - 2024-02-11

Fixed

  • Fixed the sharing of variables with role dependencies (Ansible Issue 80944) -> Support Ansible 2.15 and above.

2.2.0 - 2023-09-25

Changed

  • Changed code to reflect new name as splunk-platform-automator
  • Removed biased language and use new manager/peer terms from 9.0.0 on
  • Transfered the project to the Splunk Organization
  • Renamed the project from splunkenizer to splunk-platform-automator

Added

  • Added config setting to connect to an existing license server

Fixed

  • Fixed usage of AWS key/secret in splunk_config.yml
  • Fixed 'Unsupported parameters' for command in newer ansible versions

2.1.0 - 2022-03-05

Changed

  • Changed default image for virtualbox to almalinux/8
  • Index path settings are defined in the [default] per default. Can be disabled.
  • Index definitions at splunk_defaults.splunk_indexes must be in dictionary format
  • Changed start_ip for virtualbox, because of new restriction in Vbox 6.1.28

Added

  • Added check for acl package availability in linux
  • Added setting splunk_use_policykit to disable policykit usage
  • Added logic to setup sudo rules, if policykit is not installed
  • Added playbook to run search head cluster rolling upgrade
  • Added new splunk_hosts types: list and iter (shorter config for lots of equal nodes)
  • Added possibility to set addidional index options
  • Added test_metrics index per default
  • Added setting splunk_kv_store_engine_wiredtiger to disable config (enabled by default)
  • Added playbook to run indexer cluster rolling upgrade
  • Added playbook to cleanup splunk_backup (etc) archives
  • Added playbook to update inputs ssl certificates
  • Added playbook to call splunk rest endpoints

Fixed

  • Fixed update /etc/hosts if ip_addr not available
  • Fail on install if no splunk.secret file is found and more than one host is deployed
  • Create auth dir if needed
  • Fixed missing bundle push after shc setup
  • Fixed boolean comparison for older python versions
  • Fixed removed collections.Mapping usage with python 3.10
  • Added splunkd full service check during SHC member add. Did run into timeout when having lots of apps.
  • Updated URL for downloads of splunk install archives
  • Follow symlinks during Splunk archive extraction
  • Fixed SPLUNK_HOME ownership, when having it linked to another directory

2.0.0 - 2021-07-27

Changed

  • Never disable SELinux on Universal Forwarders
  • start_ip is now part of virtualbox section (was in general before)

Added

  • Added ability to add custom ansible variables globally and on host level
  • Set storageEngine=wiredTiger on new installs for 8.1+
  • Created Ansible inventory plugin (no more vagrant status needed to recalculate inventory)
  • Create all AWS EC2 instances at the same time. Huge time saver!
  • Better error checking in splunk_config.yml file
  • Added playbook (create_linkpage.yml) to update index.html file
  • Added disable_apparmor setting to disable AppArmor, if found
  • Added pipelining = true to ansible.cfg

Fixed

  • Usage of own certificates for single indexers

Removed

  • Ansible cannot be called from vagrant directly
  • Removed dependency to vagrant (Although vagrant does still work!)
  • Removed usage of vagrant-hostmanager plugin. This plugin can be removed.

1.3.0 - 2021-04-11

Changed

  • New default: Do not run ansible from vagrant
  • New default: disable_selinux: true (disables only, if state is 'enforcing')

Deprecated

  • Call ansible from vagrant

Added

  • Added support for SmartStore (S2)
  • Install Policy Kit Rules by enable boot-start for 8.1.1+
  • Added setting summary_replication = true
  • Added Bucket tunings for indexers
  • Do not update comment macro permissions for Splunk 8.1+
  • Added playbook to update splunk web certs
  • Support upgrades to Splunk 8.x
  • Allow to set custom config file settings (splunk_conf)
  • Option the 'os' section to disable SELinux (needs Ansible 2.7)
  • Option to allow maxVolumeDataSizeMB for volumes to be calculated from the available filesystem free space

Fixed

  • Fixed ansible temp_dir warning during splunk download
  • Fixed SHC setup for Splunk 8.1.x
  • Fixed changed status to be ok for non changing shell and command tasks
  • Fixed splunk command calls to run as user splunk instead of root
  • Do not deploy org_search_volume_indexes on single node instance
  • Fixed search head cluster setup when using own certificates
  • Enable systemd-managed by default with Splunk 7.3.x
  • Fixes broken universal forwarder install with 7.3.x
  • Fixed search head cluster setup for 7.3.x
  • Fixed hard coded ansible user home location
  • Fixed issue: Special characters in Splunk password #6
  • Added workaround for bug in vagrant-aws with vagrant 2.2.7

1.2 - 2019-06-02

Added

  • Added option to run ansible independent from vagrant.
    • Create VM first without Ansible and run playbooks in parallel on the nodes. See README.md
    • More modularization in the code and playbooks
    • Added playbook to remove the Splunk installation from nodes
    • Added playbooks to stop/start/restart splunk
    • Added playbook to uninstall the splunk software along with THP and ulimit settings
    • Added playbook to run splunk commands
    • Added playbook to upgrade the Splunk software (not cluster aware)
  • Support native systemd support introduced with Splunk Version 7.2.2
    • Added ulimit settings to native systemd service file
    • Create policy kit rule if systemd version 226 is available
    • Support polkit version 0.105 policy file.
    • Create suoders file as workaround to allow splunk user restart splunk service, if systemd version is too low.
  • Added support for Ansible versions 2.7.x and removed Ansible version check
  • Added support for Ansible versions 2.8.x
  • Update permissions of comment macro to be global
  • Moved python install on ubuntu to splunk_config file
  • Added option 'idxc_discovery_password' to setup indexer discovery
  • Support for multiple license files
  • Add option general.url_locale (ex. en-US) to be added in index.html
  • Added options to turn off login page info (ex. in AWS)
  • Added options to force inventory name set as serverName and/or host (ex. in AWS)
  • Added option to set os hostname (ex. in AWS)
  • Support for windows virtual machine setup in Virtualbox
    • Installation and configuration of windows universal forwarder (deployment server needed for now)

Fixed

  • Fixed 'vbguest' error, when using AWS only
  • Fixed some base_config installations on single node configs with additional roles
  • Disable time sync cron for AWS by default

1.1.1 - 2018-10-18

Changed

  • Optimized network config file upates for AWS instances

Fixed

  • Fixed SHC setup for Splunk 7.2
  • Fixed updating the index.html, after restart AWS instances

1.1 - 2018-10-08

Added

  • Added support for Ansible versions 2.5.x and 2.6.x
  • Added support for creating virtual machines in the Amazon Cloud (AWS)
  • Replaced internal hosts file maintenance with the vagrant hostmanager plugin
  • Improved standalone Ansible playbook usage
  • Added option to download splunk binaries from splunk.com during install
  • Added hf_host field to heavy_forwarder
  • Added feature to save a copy of the base_config apps on the Ansible host

Fixed

  • Removed locale (en_GB) in the links of node link page (index.html)
  • Fixed wrong systemd service name for universal forwarder
  • Removed Deployment Server from serverclasses whitelists
  • Do not install splk_all_forwarder_outputs on a single box, when adding LM, MC roles

1.0 - 2018-05-21

Added

  • Support hashed config values for cluster and ssl cert passwords
  • Support for using the same splunk.secret file
  • Support new password policy of Splunk 7.1 during install
  • Reworked shcluster setup. Better support for adding shc nodes.
  • Allow Splunk Version 'latest' to use the latest found splunk version in the directory
  • Support for single node system
  • Using systemd for splunk services, where available
  • Timezone taken from vagrant host per default
  • Support Ubuntu
  • Support index volume definitions
  • Support to use Ansible playbooks without vagrant
  • Support ssl for splunk web (including custom certs)
  • Support ssl for forwarder->indexer communicaion (including custom certs)

Fixed

  • Remove deployment server from the host lists generated for serverclasses
  • Allow fgdn hosts
  • Added missing org_search_volume_indexes to DS
  • Added org_all_search_base to all roles with web enabled

0.9 - 2018-02-09

Added

  • Added time sync workaround for the clock skew without virtualbox additions
  • Configuration variables can now also be set on splunk_hosts level
    • Ansible vars are configurable in the config file (ex. skip_tags)
    • Allow to install additional os packages
  • Make 'org_' for apps changeable, does set to splunk_env_name
  • Make destname for apps changeable (ex. org_site_n_indexer_base)
  • Added single indexer playbook, forwarding config to it not yet implemented
  • Added ansible.cfg to turn off deprecation warnings on ansible 2.4+
  • Add note about hostname, roles, user/pw on login page
  • Support multiple indexer clusters in org_all_forwarder_outputs (use idxc name in stanza)
  • Allow mixed splunk versions. Can be set per splunk_env or host level
  • Support single indexers
  • Support single search heads
  • Create HTML link page for all roles

Removed

  • Removed [splunk_env_name:vars] from the ansible inventory

Changed

  • Simplifyed the configuration file. Established default values for most of the settings
  • Reworked outputs and search_peer configuration

Fixed

  • not all apps on single search head are deployed from deployment server
  • add tags to the dserver and serverclass tasks
  • volume and indexes should be deployed to single search head (create serverclass)
  • org_all_forwarder_outputs should output to all indexers. does only use first cluster in multicluster config

0.8 - 2017-11-12

  • Check for site affinity in search head clusters
  • Install path /opt changeable
  • Disable indexing on CM, DS(?), Deployer (check in the cli file) -> done by the forwarder app
  • set indexes from all config file
  • org_search_volume_indexes on SH
  • Check correct usage of ansible vars in when statements
  • Set search head cluster label
  • org_cluster_search_base needs probably app_path during config on deployer, wrong path used
  • sh need splunk restart after cluster init
  • Find solution for reboot before bootstrap (if command throws error, reboot and bootstrap)
  • tries to add dservers twice on smc, dserver list must be created in a better way, indexer cluster not there
  • create hosts file out of inventory of Vagrant file (done from config file)
  • make Vagrant file dynamical
  • Set role with Vagrant var
  • make common main playbook and decide on Vagrant var (done on the config role)
  • after shcluster build, bundle push must be performed: splunk apply shcluster-bundle -target https://sh3:8089
  • uf is not connecting after install, needs reboot, how to check for it?
  • do not add own host to dservers and make the list unique
  • Multi site indexer cluster
  • single site multiside cluster
  • renamed playbooks -> ansible
  • probably need to wait until the shcluster has restarted, otherwise bundle deploy can fail
  • org_full_license_server should not be installed if role license master (check on cm)
  • org_cluster_search_base should not be installed if role is cluster master
  • ulimit settings https://docs.splunk.com/Documentation/Splunk/latest/Installation/Systemrequirements#Considerations_regarding_system-wide_resource_limits_on_.2Anix_systems
  • THP settings http://docs.splunk.com/Documentation/Splunk/latest/ReleaseNotes/SplunkandTHP
  • org_cluster_search_base misses multisite = true if multiside idx cluster
  • one more restart for UF
  • caclculate idxc_available_sites dynamically and add it to the cluster
  • calculate shc_captain dynamically
  • Change config to be able to have more than one cluster
  • do not provide base_config apps, user must download
  • Support multiple cluster masters in org_cluster_search_base (use idxc name in stanza)
  • hashed password check must be changed to check inside the cm stanza in org_cluster_search_base on cm
  • Add cm as search peer in SMC
  • Support single search heads
  • don't install relevant baseconfig, if clustermaster, deployment server, monitoring console is not there
  • check if full_lic app is created, if license master is not already installed, works :-)
  • check if deployment client app is installed, if deployment server is not yet installed, works :-)
  • add forward app, if deployment server is missing
  • add license client app, if deployment server is missing
  • allow relative pathes for software an baseconfigs
  • added org_all_search_base
  • fix dserver list creation fail, if group was missing
  • Output error if license file not there
  • Added fix for clock skew on Linux in Virtualbox: https://oitibs.com/fix-virtualbox-guest-time-skew (not working, either)
  • dserver list does not exclude UF
  • Install forward output, if deployer is ds
  • Install forward output, if cm is ds
  • Create list of license clients in serverclass dynamically
  • added new role for heavy forwarder
  • cleanup ansible tags
  • Calculate /etc/hosts from inventory if possible otherwise from splunk_config
  • Single SH does not get license client app -> need to be added automatically in server class
  • reload deploy server, if serverclass changes
  • fix serverclass whitelisting
  • dynamic ip addresses
  • use a start_ip for dynamic ips
  • update serverclass everytime a server is created
  • Check host to be available before adding dserver
  • set site0 to cluster masters in the example configs (no site affinity)
  • Allow to turn on virtualbox addon installation inside config (disable clock skew fix)
  • disable first login page