ManageIQ Fine-3 HA in VMware

fine
ui

#1

Hi all ,

I`ll appreciate if someone can point me to some good documentation or helpful page to setup HA ManageIQ Fine-3 in Vmware ENV .

I have searched around Documentation of ManageIQ but doesn`t found . I have followed some CloudForms tutorials but still no success . what I would like to setup is 2 Web Appliance for 1 Regio in HA mode .

followed @carbonin in talk.manageiq.org . read all comments all topics but still no success .
Installation image is downloaded from http://releases.manageiq.org/manageiq-vsphere-fine-3.ova

I have tried PROBLEM: With Generate Custom Encryption Key .

as of https://access.redhat.com/documentation/en-us/red_hat_cloudforms/4.2/html/general_configuration/configuration#database-regions-and-replication
and etc but I cant get 2 appliance in replication .

I`ll post some helpfull docs maybe someone can get help from here
https://access.redhat.com/documentation/en-us/red_hat_cloudforms/4.2/html/general_configuration/configuration#regions



but quite old .

Regards
Karen


#3

@kalinux Unfortunately, there are two different features in ManageIQ that can be described as database replication.

  1. Multiple regions can replicate data to a central region, which can be used as a single pane of glass for reporting and also some admin tasks. This type of replication is configured using the ManageIQ web interface and is usually what people are talking about when they say “Replication”. This type of replication is implemented using logical replication (pglogical) of particular tables and is not suitable for highly available environments.
  2. HA replication is implemented using PG physical streaming replication and is configured using the appliance console.

Generally I would suggest not co-locating a running instance of the application with a database in an HA scenario. ManageIQ appliances automatically configure the database on first boot, so there are a few steps to go through before you will be able to configure the database separately.

The following will stop the application and the database, remove the database that was configured on first boot and get you into a situation where you should be able to follow the documentation here.

systemctl stop evmserverd
systemctl disable evmserverd
systemctl stop $APPLIANCE_PG_SERVICE
systemctl disable $APPLIANCE_PG_SERVICE
vmdb
rm -f REGION config/database.yml
rm -rf $APPLIANCE_PG_DATA/*

At this point you will probably want to create a template from this machine and use it to deploy the rest of the environment. The documentation I linked should get you through setting up the database cluster.

My suggested deployment would be at least two database appliances (preferably 3) and then however many application appliances you need. Also, if you think you need more storage for the database, feel free to resize the existing volume on the appliance. We ship with a logical volume configured for the database so you should be able to add more physical volumes to that to extend the size.

Hope this helps,
Nick


Manageiq appliance_console “configure database”->“cretae key”failed!
#4

Awesome ! thanks @carbonin . Im going to give a try above described steps , and Ill send feedback !

Regards
Kalinux


#5

Hi @carbonin .

Above steps worked very well . Thank you very much ! .

But now lets move little bit further and I have test setup with 5 Appliance . which are .
app-01
app-02
db-01 (primary)
db-02(standby)
db-03(standby)

between app-01 and app-02 I`ll setup LoadBalancer . but is there any suggestion from ManageIQ side about loadbalancer ? My plan is to setup HA-proxy or usual NGINX .

One more not clear part in this picture , what if my db-01(primary) goes down ? db-02 and db-03 has automatic failover , but how does app-01 and app-02 switch to db-02 or db-03 ?

Thank you!
Regards
Karlinux


#6

There is some documentation around configuring HAProxy here, but I’m not very familiar with that configuration.

For the app servers to recognize and deal with a failover event we have a monitor process that runs on them.
https://github.com/ManageIQ/manageiq_docs/blob/master/doc-High_Availability_Guide/topics/Database_Failover.adoc should give more information about that.