Provisioning Configuration

Provisioning Settings

Since version 1.1 replication-manager can use an agent-based cluster provisioning with the OpenSVC provisioning framework.

Starting with version 2.0 provisioning is packaged in a separate binary called replication-manager-pro

./replication-manager-pro monitor

The following software services can be provisioned:

Item	Docker	Package
MariaDB	Y	Y
MySQL	Y	Y
Percona	Y	Y
MaxScale	Y	Y
ProxySQL	Y	N
HaProxy	Y	N
Sphinx	Y	N
Consul	N	N

replication-manager is using a secure client API to the OpenSVC collector. This collector is used for posting actions to a cluster of agents, fetch cluster nodes information and uploading his own set of playbooks for provisioning.

The Signal18.io SAS collector can be used for faster testing or not to have to maintain an extra piece of infrastructure, if not possible, an evaluation version of the collector needs to be installed, or it can be installed co-located to replication-manager or on a separate machine.

replication-manager talks to the default SAS collector or it can be setup to talk to on premise collector via the following parameters:

`prov-orchestrator` (2.1)

Item	Value
Description	Orchestration type
Type	String
Values	onpremise	opensvc	kube	slapos	local
Default	onpremise	opensvc	kube	slapos	local for pro release or onpremise	local for osc release
Example	"opensvc"

`opensvc-host` (1.1)

Item	Value
Description	Address of the OpenSVC collector
Type	String
Default	"ci.signal18.io:9443"
Example	"127.0.0.1:443"

`opensvc-admin-user` (1.1)

Item	Value
Description	Admin credential of the OpenSVC collector
Type	String
Default	"[email protected]:opensvc"
Example	"[email protected]:secret"

For on premise collector it is needed to make a first registration on the collector to create a regular user, group, roles and compliances.

replication-manager-pro monitor --opensvc-register

To use the SAS monitor you need to login to signal18.io and request for your evaluation or licence account.yaml file. Copy it into the share/opensvc directory of replication-manager-pro

Services Options

micro-services are isolating a collection of resources, each service can than be used for easy management, HA policy like DR plan in agent 1.8 and placement over a cluster on agent version 1.9

An existing disk device, if none a loopback device to store service data
An existing disk device, if none a loopback device to store the docker data
A file system type zfs|ext4|xfs|hfs|aufs
A file system pool type lvm|zpool|none
An IP address that needs to be unused in the network

Resources choice is uniform over a full cluster.

`prov-db-service-type` (1.1)

Item	Value
Description	Database type of Micro-Services deployment
Type	Enum
Values	Docker,Package
Example	"Docker"

Type of Micro-Services can be docker or package not that if package it need the package install on the agent as replication-manager will only call the binary for bootstrapping and expect it to be present on the agent.

`prov-db-agents` (1.1)

Item	Value
Description	List of agents for Database Micro-Services placement
Type	List
Values	Docker,Package
Example	"Docker"

The agent names can be found in the web interface under the agents tab.

Disk

`prov-db-docker-img` (1.1)

Item	Value
Description	The database docker image to deploy
Type	String
Example	"mariadb:latest"

`prov-db-disk-fs` (1.1)

Item	Value
Description	Database type of FS deployment
Type	Enum
Values	xfs,ext4,zfs,ufs
Example	"zfs"

File system many drivers are available we do test xfs ext4 zfs . Many other drivers like ceph or drbd would need additional testing to be used as extra options may need to be added.

OpenSVC Agent drivers can provision disk resources on many SAN arrays and Cloud API, contact support if you need custom type of disk provisioning for your architecture.

`prov-db-disk-pool` (1.1)

Item	Value
Description	Database disk pool type for Micro Services deployment
Type	Enum
Values	none,zpool,lvm
Example	"zpool"

`prov-db-disk-type` (1.1)

Item	Value
Description	Database disk pool type for Micro Services deployment
Type	Enum
Values	loopback,physical,pool,directory
Example	"loopback"

When loopback instead of a real device the FS path is needed instead of device path

`prov-db-disk-device` (1.1)

Item	Value
Description	Database disk device path for Micro Services deployment
Type	String
Example	"/srv"

Depends on prov-db-disk-type

physical: define the device /dev/XXXXXXXX pool: define the pool name loopback: define the path to create the loopback file directory: define the path to create the service_name directory

`prov-db-disk-size` (1.1)

Item	Value
Description	Databse disk size in g for micro service VM .
Type	String
Default	"20g"
Example	"20g"

`prov-db-disk-snapshot` (2.0)

Item	Value
Description	Take daily snapshot.
Type	boolean
Default	false

`prov-db-disk-snapshot-keep` (2.0)

Item	Value
Description	Keep this number of snapshot.
Type	int
Default	7

network

Network please check availability of the ip before using them , also some opensvc deployemetn can manage range of dhcp ip and DNS entries

`prov-db-net-iface` (1.1)

Item	Value
Description	Database ethernet device.
Type	String
Example	"br0"

`prov0-db-net-gateway` (1.1)

Item	Value
Description	Database network gateway.
Type	String
Example	"192.168.1.254"

`prov-db-net-mask` (1.1)

Item	Value
Description	Database eth device
Type	String
Example	"255.255.255.0"

Resource Usage

Database bootstrap is deploying some database configurations files that are auto adapted to following cluster parameters and to tags:

`prov-db-memory` (1.1)

Item	Value
Description	Database memory in M for micro service.
Type	String
Example	"256"

`prov-db-disk-iops` (1.1)

Item	Value
Description	Database Rnd IO/s in for micro service.
Type	String
Default	"300"
Example	"300"

`prov-db-cpu-cores` (2.0)

Item	Value
Description	Database number of cores for micro service.
Type	String
Default	"1"
Example	"4"

Database tags:

`prov-db-tags (1.1)`

Item	Value
Description	Database tags for compliance configuration.
Type	String
Example	"innodb,noquerycache,threadpool,logslow"

Storage:

innodb, myrocks, tokudb, spider, sphinx

Logs:

logaudit, logslow, logsqlerrors, loggeneral, logpfs, loguserstats

Features:

compress, threadpool, ssl, lowercasetable, smallredolog,sqlmodeunstrict, sqlmodeoracle
 noquerycache,  nodurable, nodoublewrite,  noautocommit, noodirect

Replication:

multidomains, nologslaveupdates, mysqlgtid, wsrep, semisync

Provisioning

Micro-services placement will follow a round robin mode against the agents listed for a service.

bootstrap and unprovision commands can be found in the web interface.

The client can also be used to fully provision a cluster defined in the configuration.

replication-manager-cli bootstrap  --cluster=cluster_haproxy_masterslave --with-provisioning
Provisioning done

Provisioning Configuration

Provisioning Settings

prov-orchestrator (2.1)

opensvc-host (1.1)

opensvc-admin-user (1.1)

Services Options

prov-db-service-type (1.1)

prov-db-agents (1.1)

Disk

prov-db-docker-img (1.1)

prov-db-disk-fs (1.1)

prov-db-disk-pool (1.1)

prov-db-disk-type (1.1)

prov-db-disk-device (1.1)

prov-db-disk-size (1.1)

prov-db-disk-snapshot (2.0)

prov-db-disk-snapshot-keep (2.0)