mirror of
https://github.com/apache/cloudstack.git
synced 2025-10-26 08:42:29 +01:00
30175d6879
The new CA framework introduced basic support for comma-separated list of management servers for agent, which makes an external LB unnecessary. This extends that feature to implement LB sorting algorithms that sorts the management server list before they are sent to the agents. This adds a central intelligence in the management server and adds additional enhancements to Agent class to be algorithm aware and have a background mechanism to check/fallback to preferred management server (assumed as the first in the list). This is support for any indirect agent such as the KVM, CPVM and SSVM agent, and would provide support for management server host migration during upgrade (when instead of in-place, new hosts are used to setup new mgmt server). This FR introduces two new global settings: - `indirect.agent.lb.algorithm`: The algorithm for the indirect agent LB. - `indirect.agent.lb.check.interval`: The preferred host check interval for the agent's background task that checks and switches to agent's preferred host. The indirect.agent.lb.algorithm supports following algorithm options: - static: use the list as provided. - roundrobin: evenly spreads hosts across management servers based on host's id. - shuffle: (pseudo) randomly sorts the list (not recommended for production). Any changes to the global settings - `indirect.agent.lb.algorithm` and `host` does not require restarting of the mangement server(s) and the agents. A message bus based system dynamically reacts to change in these global settings and propagates them to all connected agents. Comma-separated management server list is propagated to agents on following cases: - Addition of a host (including ssvm, cpvm systevms). - Connection or reconnection by the agents to a management server. - After admin changes the 'host' and/or the 'indirect.agent.lb.algorithm' global settings. On the agent side, the 'host' setting is saved in its properties file as: `host=<comma separated addresses>@<algorithm name>`. First the agent connects to the management server and sends its current management server list, which is compared by the management server and in case of failure a new/update list is sent for the agent to persist. From the agent's perspective, the first address in the propagated list will be considered the preferred host. A new background task can be activated by configuring the `indirect.agent.lb.check.interval` which is a cluster level global setting from CloudStack and admins can also override this by configuring the 'host.lb.check.interval' in the `agent.properties` file. Every time agent gets a ms-host list and the algorithm, the host specific background check interval is also sent and it dynamically reconfigures the background task without need to restart agents. Note: The 'static' and 'roundrobin' algorithms, strictly checks for the order as expected by them, however, the 'shuffle' algorithm just checks for content and not the order of the comma separate ms host addresses. Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>
210 lines
8.7 KiB
Properties
210 lines
8.7 KiB
Properties
# Licensed to the Apache Software Foundation (ASF) under one
|
|
# or more contributor license agreements. See the NOTICE file
|
|
# distributed with this work for additional information
|
|
# regarding copyright ownership. The ASF licenses this file
|
|
# to you under the Apache License, Version 2.0 (the
|
|
# "License"); you may not use this file except in compliance
|
|
# with the License. You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing,
|
|
# software distributed under the License is distributed on an
|
|
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
|
|
# KIND, either express or implied. See the License for the
|
|
# specific language governing permissions and limitations
|
|
# under the License.
|
|
|
|
# Sample configuration file for CloudStack agent
|
|
|
|
# The GUID to identify the agent with, this is mandatory!
|
|
# Generate with "uuidgen"
|
|
guid=
|
|
|
|
#resource= the java class, which agent load to execute
|
|
resource=com.cloud.hypervisor.kvm.resource.LibvirtComputingResource
|
|
|
|
#workers= number of threads running in agent
|
|
workers=5
|
|
|
|
#host= The IP address of management server
|
|
host=localhost
|
|
|
|
# The time interval in seconds after which agent will check if connected host
|
|
# is the preferred host (the first host in the comma-separated list is preferred
|
|
# one) and will attempt to reconnect to the preferred host when it's connected
|
|
# to one of the secondary/backup hosts. The timer task is scheduled after agent
|
|
# connects to a management server. On connection, it receives admin configured
|
|
# cluster-level 'indirect.agent.lb.check.interval' setting that will be used by
|
|
# the agent as the preferred host check interval however the following setting
|
|
# if defined overrides the received value. The value 0 and lb algorithm 'shuffle'
|
|
# disables this background task.
|
|
#host.lb.check.interval=0
|
|
|
|
#port = The port management server listening on, default is 8250
|
|
port=8250
|
|
|
|
#cluster= The cluster which the agent belongs to
|
|
cluster=default
|
|
|
|
#pod= The pod which the agent belongs to
|
|
pod=default
|
|
|
|
#zone= The zone which the agent belongs to
|
|
zone=default
|
|
|
|
#public.network.device= the public nic device
|
|
# if this is commented, it is autodetected on service startup
|
|
# public.network.device=cloudbr0
|
|
|
|
#private.network.device= the private nic device
|
|
# if this is commented, it is autodetected on service startup
|
|
# private.network.device=cloudbr1
|
|
|
|
#guest.network.device= the guest nic device
|
|
# if this is commented, the private nic device will be used
|
|
|
|
# local storage path, by default, it's /var/lib/libvirt/images/
|
|
#local.storage.path=/var/lib/libvirt/images/
|
|
|
|
# Qemu socket path, directory where Qemu sockets are placed.
|
|
# These sockets are for the Qemu Guest Agent and SSVM privisioning
|
|
# Make sure that AppArmor or SELinux allow libvirt to write there
|
|
#qemu.sockets.path=/var/lib/libvirt/qemu
|
|
|
|
# The UUID for the local storage pool, this is mandatory!
|
|
# Generate with "uuidgen"
|
|
local.storage.uuid=
|
|
|
|
# Location for KVM scripts
|
|
domr.scripts.dir=scripts/network/domr/kvm
|
|
|
|
# the timeout for time-consuming operations, such as create/copy snapshot
|
|
#cmds.timeout=7200
|
|
|
|
# set the vm migrate speed, by default, it will try to guess the speed of the guest network
|
|
# In MegaBytes per second
|
|
#vm.migrate.speed=0
|
|
|
|
# set target downtime at end of livemigration, the 'hiccup' for final copy. Higher numbers
|
|
# make livemigration easier, lower numbers may cause migration to never complete. Less than 1
|
|
# means hypervisor default (20ms).
|
|
#vm.migrate.downtime=0
|
|
|
|
# Busy VMs may never finish migrating, depending on environment. When its available, we will
|
|
# want to add support for autoconvergence migration flag which should fix this. Set an upper
|
|
# limit in milliseconds for how long live migration should wait, at which point VM is paused and
|
|
# migration will finish quickly. Less than 1 means disabled.
|
|
#vm.migrate.pauseafter=0
|
|
|
|
# set the type of bridge used on the hypervisor, this defines what commands the resource
|
|
# will use to setup networking. Currently supported NATIVE, OPENVSWITCH
|
|
#network.bridge.type=native
|
|
|
|
# set the driver used to plug and unplug nics from the bridges
|
|
# a sensible default will be selected based on the network.bridge.type but can
|
|
# be overridden here.
|
|
# native = com.cloud.hypervisor.kvm.resource.BridgeVifDriver
|
|
# openvswitch = com.cloud.hypervisor.kvm.resource.OvsVifDriver
|
|
#libvirt.vif.driver=com.cloud.hypervisor.kvm.resource.BridgeVifDriver
|
|
|
|
# set the hypervisor type, values are: kvm, lxc
|
|
hypervisor.type=kvm
|
|
|
|
# set the hypervisor URI. Usually there is no need for changing this
|
|
# For KVM: qemu:///system
|
|
# For LXC: lxc:///
|
|
# hypervisor.uri=qemu:///system
|
|
|
|
# settings to enable direct networking in libvirt, should not be used
|
|
# on hosts that run system vms, values for mode are: private, bridge, vepa
|
|
# libvirt.vif.driver=com.cloud.hypervisor.kvm.resource.DirectVifDriver
|
|
# network.direct.source.mode=private
|
|
# network.direct.device=eth0
|
|
|
|
# setting to enable the cpu model to kvm guest globally.
|
|
# three option:custom,host-model and host-passthrough.
|
|
# custom - user custom the CPU model which specified by guest.cpu.model.
|
|
# host-model - identify the named CPU model which most closely matches the host,
|
|
# and then request additional CPU flags to complete the match. This should give
|
|
# close to maximum functionality/performance, which maintaining good
|
|
# reliability/compatibility if the guest is migrated to another host with slightly different host CPUs.
|
|
# host-passthrough - tell KVM to passthrough the host CPU with no modifications.
|
|
# The difference to host-model, instead of just matching feature flags,
|
|
# every last detail of the host CPU is matched. This gives absolutely best performance,
|
|
# and can be important to some apps which check low level CPU details,
|
|
# but it comes at a cost wrt migration. The guest can only be migrated to
|
|
# an exactly matching host CPU.
|
|
#
|
|
# guest.cpu.mode=custom|host-model|host-passthrough
|
|
# This param is only valid if guest.cpu.mode=custom,
|
|
# for examples:"Conroe" "Penryn", "Nehalem", "Westmere", "pentiumpro" and so
|
|
# on,run virsh capabilities for more details.
|
|
# guest.cpu.model=
|
|
#
|
|
# This param will require CPU features on the <cpu> section
|
|
# guest.cpu.features=vmx vme
|
|
#
|
|
# vm.memballoon.disable=true
|
|
# Disable memory ballooning on vm guests for overcommit, by default overcommit
|
|
# feature enables balloon and sets currentMemory to a minimum value.
|
|
#
|
|
# vm.diskactivity.checkenabled=false
|
|
# Set to true to check disk activity on VM's disks before starting a VM. This only applies
|
|
# to QCOW2 files, and ensures that there is no other running instance accessing
|
|
# the file before starting. It works by checking the modify time against the current time,
|
|
# so care must be taken to ensure the cluster has time synced, otherwise VMs may fail to start.
|
|
#
|
|
# vm.diskactivity.checktimeout_s=120
|
|
# Timeout for giving up on waiting for VM's disk files to become inactive. Hitting
|
|
# this timeout will result in failure to start VM.
|
|
#
|
|
# vm.diskactivity.inactivetime_ms=30000
|
|
# This is the length of time that the disk needs to be inactive in order to pass the check.
|
|
# This means current time minus mtime of disk file needs to be greater than this number.
|
|
# It also has the side effect of setting the minimum threshold between a stop and start of
|
|
# a given VM.
|
|
#
|
|
# kvmclock.disable=false
|
|
# Some newer linux kernels are incapable of reliably migrating vms with kvmclock
|
|
# This is a workaround for the bug, admin can set this to true per-host
|
|
#
|
|
# vm.rng.enable=false
|
|
# This enabled the VirtIO Random Number Generator device for guests.
|
|
#
|
|
# vm.rng.model=random
|
|
# The model of VirtIO Random Number Generator (RNG) to present to the Guest.
|
|
# Currently only 'random' is supported.
|
|
#
|
|
# vm.rng.path=/dev/random
|
|
# Local Random Number Device Generator to use for VirtIO RNG for Guests.
|
|
# This is usually /dev/random, but per platform this might be different
|
|
#
|
|
# vm.rng.rate.bytes=2048
|
|
# The amount of bytes the Guest may request/obtain from the RNG in the period
|
|
# specified below.
|
|
#
|
|
# vm.rng.rate.period=1000
|
|
# The number of milliseconds in which the guest is allowed to obtain the bytes
|
|
# specified above.
|
|
#
|
|
# router.aggregation.command.each.timeout=600
|
|
# timeout value for aggregation commands send to virtual router
|
|
#
|
|
# host.overcommit.mem.mb = 0
|
|
# allows to increase amount of ram available on host virtually to utilize Zswap, KSM features
|
|
# and modern fast SSD/3D XPoint devices. Specified amount of MBs is added to the memory agent
|
|
# reports to the Management Server
|
|
# Default: 0
|
|
#
|
|
# vm.watchdog.model=i6300esb
|
|
# The model of Watchdog timer to present to the Guest
|
|
# For all models refer to the libvirt documentation.
|
|
# Recommend value is: i6300esb
|
|
#
|
|
# vm.watchdog.action=none
|
|
# Action to take when the Guest/Instance is no longer notifiying the Watchdog
|
|
# timer.
|
|
# For all actions refer to the libvirt documentation.
|
|
# Recommended values are: none, reset and poweroff.
|