Archive

Posts Tagged ‘vCloud Suite’

Configuring vRealize Log Insight Agent on MS SQL Server 2012

There’s a lot of information available about the SQL Server Content Pack for Log Insight, but actually  trying to use it reveals a lot of gaps.  So, here’s what I’ve done to get my SQL Server 2012 R2 instances to correctly report their information to vRealize Log Insight and show up in the Content Pack filters.

To ensure the correct log directories are used and allow reuse of filelog sections, I suggest creating an agent group for each SQL Server.  I’m also a fan of the server-side agent configuration, so if the agent gets totally removed and reinstalled, it’ll acquire the correct configuration again.

Steps:

  1. Install the Log Insight Agent for Windows on the SQL Server machine, be sure to use the FQDN of the Log Insight VIP
  2. While logged on to the SQL Server, locate the Logs folder for each SQL server instance. Note that this is not the database transaction logs folder – if it contains *.ldf files, it’s the wrong folder.   Specifically, you want the folder that contains the ERRORLOG file.  To be certain follow these steps:
    • Launch SQL Server Configuration Manager
    • Select “SQL Server Services” from the left pane
    • Right-click “SQL Server (INSTANCENAME)” from the right page, choose “Properties”.
    • Select the “Startup Parameters” tab
    • In the Existing Parameters box, make a note of the parameter that starts with “-e”
    • The filepath in that parameter (minus “ERRORLOG”) is the folder we need.
    • DO NOT CHANGE ANY VALUES, Cancel Properties, Close Configuration Manager

Startup Params in ConfigurationManager

  1. Logon to Log Insight as an administrator
  2. Check the Content Package area to be sure the “Microsoft – SQL Server” Content Pack is installed
  3. In the next steps, we’ll configure the agent from the server-side. It is possible to do from the client-side.
  4. In the Administration|Agents section of Log Insight select Microsoft – SQL Server from Available Templates drop-down list
  5. Click the “Copy Template” button. In the Copy Agent Group box that appears, change the name to “Microsoft – SQL Server – SERVERNAME”.  (replace SERVERNAME with the actual server’s name).  Click “Copy”.
  6. Now, the dropdown list should indicate that you are editing the new Agent Group
  7. In the Filter box, edit the parameters to create a filter that identifies only the desired server. For example “Hostname | starts with | SERVERNAME”
  8. In the Agent Configuration section
    • Update the section header from “[filelog|MSSQL]” to [filelog|MSSQL-SERVERNAME-INSTANCENAME]” replace SERVERNAME and INSTANCENAME with the correct, actual values. Be sure to keep “MSSQL” as the prefix, since the dashboard elements key on that.
    • Update the directory value to where the ERRORLOG file was found
    • Update the exclude value to read *.trc;*.xel;*.mdmp
    • Example:

[filelog|MSSQL-COREDB1-MSQLSERVER]
; IMPORTANT: Change the directory as per the environment
directory=C:\MSSQL\DATA\MSSQL12.MSSQLSERVER\MSSQL\Log
tags={“ms_product”:”mssql”}
charset=UTF-16LE
exclude=*.trc;*.xel;*.mdmp

  1. If there are multiple instances of SQL server on the host server, copy the entire section and edit the section header with the Instance Name. Example of configuration for multiple instances:

[filelog|MSSQL-COREDB1-MSSQLSERVER]
; IMPORTANT: Change the directory as per the environment
directory=C:\MSSQL\DATA\MSSQL12.MSSQLSERVER\MSSQL\Log
tags={“ms_product”:”mssql”}
charset=UTF-16LE
exclude=*.trc;*.xel;*.mdmp

[filelog|MSSQL-COREDB1-WORKLOAD]
; IMPORTANT: Change the directory as per the environment
directory=C:\MSSQL\DATA\MSSQL12.WORKLOAD\MSSQL\Log
tags={“ms_product”:”mssql”}
charset=UTF-16LE
exclude=*.trc;*.xel;*.mdmp

  1. Click Save New Group.
  2. On the SQL Server, restart the “VMware vRealize Log Insight Agent”
  3. Navigate to %ALLUSERSPROFILE%\VMware\Log Insight Agent and open liagent-effective.ini with notepad. Check that the sections added above appear in this file.
    • If they do not, you may have to adjust the filters on the Agent Group
Advertisements

Fix – Unable to import vCAC/vRA certificates into Orchestrator

07/17/2015 Comments off

Problem:

While in the vRealize Orchestrator Client you find that the Library/Configuration/SSL Trust Manager/”Import a certificate from URL” workflow returns an error reading “InternalError: handshake alert: unrecognized_name” when provided. The URL the resolves to the Load-Balancer VIP for the vCAC/vRA appliances.

 

Background:

Signed SSL certificate installed on vCAC/vRA Appliance, SSL Passthrough on NSX/vCNX Load-Balancer, vCAC/vRA Settings/Hostname set to resolve to VIP, matching SSL cert.

 

Fix:

  1. SSH into the vCAC Appliance as root
  2. Backup /etc/apache2/vhosts.d/vcac.conf to vcac.conf.bak
  3. Use vi to edit /etc/apache2/vhosts.d/vcac.conf
  4. Scroll down to  <virtualHost _default_:443>
  5. Add these lines

    ServerName fqdn.of.appliance.node

    ServerAlias: load.balancer.name

  6. Scroll further to ensure these params aren’t listed elsewhere, remove or revise if so.
  7. save the file and exit vi
  8. restart the vCAC/vRA services

Automating NSX Security Groups with vCAC/vRA – Part 2

***UPDATE***  The download link is currently broken.   I seem to have lost the file, will fix the link as soon as I find or recreate it.  Sorry about that. 😦

In part 1 of this series, we created a list of security groups and displayed that list to users during the request.  In this post, we want to enhance that functionality by adding these features.

  • Creation of Security Groups and inclusion in Dropdown lists
  • Add a VM to a Security Group post-provisioning
  • Import existing Security Groups into vRA inventory
  • Add a Security Group to a Dropdown list

We’re going to do that by importing a vCO package with some new workflows and actions, then link up the workflows to Advanced Services and Resource Actions.  As before, we’ll require the NSX and vCAC/vRA plugins for Orchestrator.

Preparation

  • Complete the creation of the VCNS.SecurityGroup.Names.production property dictionary and valuelist attribute from Part 1.  We’re going to reuse those items. so make a note of the exact name of the property dictionary and the valuelist attribute.  In my case, I’ve named the Property Dictionary VCNS.SecurityGroup.Names.production and also named the valuelist attribute for it VCNS.SecurityGroup.Names.production
  • Make sure vRA Advanced Services Server Configuration is complete and test the connection to the Orchestrator server.  The default, built-in VCO is fine.
  • Login to vCO client as a vCO Admin. Set the mode to “Design” and navigate to the Inventory tab.  Make sure that you have a connection listed under “vCAC Infrastructure Administration” and a connection listed under “NSX”.

 

Confirm that you have the necessary connections on the inventory tab

Confirm that you have the necessary connections on the inventory tab

Get the Package

I’ve put together a handful of workflows and actions that use or expand the NSX plugin to provide information of functionallity back to vRA.  Where possible, I reused existing library workflows, but in some cases, I had to use the API to create a REST call and consume that in an action.

By downloading any code, package or file, you acknowledge that:

There is no explicit or implied warranty or support for the code.  Neither Brian Ragazzi, his employer nor anyone else is responsible for any problems, errors, omissions, unexpected behavior, breakage, trauma, outage, fatigue, lost time, lost work or incontinence that may occur as a result of using the code or package.

Download the zip file.  It contains the package and a couple of images that can be used for the advanced services

Import the Package

  1. Extract the zip file
  2. In the vCO Client, navigate to the packages tab.
  3. Click the “import package” button and select the extracted .package file
  4. On the Package Import Information step, click “Import

    Package Import Information

    Package Import Information

  5. On the Import package… step, check the “Select/Deselect all” box to check all of the items.  Please note the server path, these should not be duplicates of anything else you have in your vCO inventory (unless you’ve already imported this package previously).  Click “Import Selected elements”.

    Select all items

    Select all items

  6. Review the workflows and actions added to your inventory.

Configure Advanced Services – Create Security Group

This service enables the user to create a new NSX Security Group and automatically adds its name to the appropriate dropdown list of security groups.  It can be added once for each different list of security groups.  You’ll need to know the exact name of the Property Dictionary and valuelist attribute you created in Part 1.

NSX Management Services

NSX Management Services

NSX Security Groups in vRA Items

NSX Security Groups in vRA Items

  1. While logged into vRA as a service architect, navigate to the Advanced Services tab, click “Custom Resources
  2. We need to make vRA aware of NSX Security Groups.  Click the Add button.
  3. In the Orchestrator Type field, enter “NSX:SecurityGroup“; for the Name, I suggest “NSX Security Group“, click Next

    Add NSX Security Group as Custom Resource

    Add NSX Security Group as Custom Resource

  4. On the details form, we’re not going to make any changes, but if you wanted to hide certain properties, you could here.  Click “Add“.
  5. Click “Service Blueprints”, then the “Add” button.
  6. On the Workflow tab, select the AddNewSecurityGrouptoDropdown workflow, click next.

    Select "AddNewSecurityGrouptoDropdown" workflow

    Select “AddNewSecurityGrouptoDropdown” workflow

  7. On the Details tab, set the name to something like “Create new Production NSX Security Group“, because we’re going to create the security group and add its name to the “production”dropdown list.  Click Next.

    Set the Service Item Name

    Set the Service Item Name

  8. On the Blueprint  Form tab, under the “Step” Form page (default), mouseover the text field labelled “Name of Custom Property Dictionary in vCAC/vRA”.  Click the pencil “edit” icon when it appears.

    Edit the Form Fields

    Edit the Form Fields

  9. Click the Constraints tab of the “Edit Form Field” window.  On the Value field, select “Constant” and enter “VCNS.SecurityGroup.Names.Production” (or whatever suffix you used) for the Property Dictionary.  Set the Visible value to “No” so it doesn’t show up. Click Submit on the Edit Form Field window.

    Set the name of the Property Dictionary to be updated

    Set the name of the Property Dictionary to be updated

  10. Using the same method, set the Name of the Attribute appropriately and its visibility to no
  11. Edit the “Value to be appended to the ValueList attribute” field.  Set the label to “New Security Group Name“.  Do not set a value or make this one invisible, we need the user to enter a value, submit to save.
  12. Edit the vCACIaaSHost field – using the Constraints tab again– when setting the value, choose constant, then click Add by the green plus, to display a treeview, where you can choose your connection to the IaaS Server.  Visible: No, submit to save.

    Select connection to IaaS host

    Select connection to IaaS host

  13. Edit the “NSX endpoint” field in the same way, selecting the NSX connection.
  14. When done, all fields except “New Security Group Name” will have a value.  Click Next.
  15. On the Provisioned Resource tab, select “securityGroup [NSX Security Group]“.  Click Add to save the service blueprint.
  16. Repeat steps 5-15 for any other dropdown lists containing security groups; say “Non-Production” for instance
  17. Highlight the Service Blueprint and click “Publish” to make the blueprint available for entitlements
  18. Navigate to Administration, Services.
  19. Add a new Service named “NSX Management” (for example) – I included a nifty image in the zip file
  20. Under Catalog Items, click the “Create new Production NSX Security Group” item to edit it.
  21. The Catalog item should inherit the Security Group icon from vCO, set its Service to “NSX Management”,click update to save.
  22. Create or Edit an entitlement to include the new Service and/or catalog item.
  23. Try it out, confirm that the Security Group was created in NSX, is visible in vCAC items and it name was added to the Property Dictionary

Configure Advanced Services – Import Security Group

This service allows you to make existing security groups visible as items in the vCAC Items view.  Once this is done, we’ll add actions that allow you to add the security group to a dropdown list.

  1. Click “Service Blueprints”, then the “Add” button.
  2. Select the “GetNSXSecurityGroup” workflow, click Next
  3. On the details tab, set the name to “Import NSX Security Group“, click Next
  4. On the Blueprint form, set the “connection” to the NSX connection in vCO, then hide the field.  Security Group Name will be a dropdown list of existing NSX Security Groups for the user to choose from. Click Next
  5. On the Provisioned Resource tab, select “securityGroup [NSX Security Group]“.  Click Add to save the service blueprint.
  6. Just as before, publish the service blueprint, add it to a service and an entitlement.

 Configure Advanced Services – Add Security Group to Dropdown list

With this service, we’ll let the user add the name of an existing Security Group to a drop down list.  Unlike the first two, this is implemented as a Resource Action, meaning it’ll be executed against an existing item (a Security Group in this case)

  1. Under Resource Actions, click “Add”
  2. For the Workflow, select the “AddExistingSecurityGrouptoDropdown“, click Next
  3. On the “Input Resource” tab, keep NSX Security Group, click Next
  4. On the Details tab, set the Name to “Add Security Group to Production list” or similar, set the description, leave the Type options unchecked.  click Next

    Set Action Name and Description

    Set Action Name and Description

  5. On the Form tab, just like the first service blueprint, set the Property Dictionary and Attribute names as appropriate.  VCNS.SecurityGroup.Names.production in my example, set visible to no on both.
  6. Again, we’ll set the vCACIaaSHost to the connection to the Server and hide the field
  7. Click Add to save the action.
  8. Repeat steps 1-6 for each security group dropdown list (say “non-production” for instance)
  9. Publish the action and add it to an entitlement

    Add Action to Entitlement

    Add Action to Entitlement

  10. Test by navigating toNSX under Items, highlight a Security group and Select “Add Security Group to…” from the Actions menu.

    Yay! A Resource Action

    Yay! A Resource Action

 Configure Advanced Services – Add VM to a Security Group

This service lets you add a provisioned VM to additional Security Groups.  So, at provisioning-time, the VM is added to the Security Group selected by the user, but we may need to refine the security by adding that VM to additional Security Groups.

  1. Under Resource Actions, click “Add
  2. For the Workflow, select the “AddVMtoSecurityGroup“, click Next
  3. On the “Input Resource” tab, keep IaaS VC VirtualMachine, click Next
  4. On the Details tab, set the name to “Add VM to a Security Group“, click Next
  5. On the Form tab, set the connection Value to the NSX connection.
  6. Leave the NSX Security Group field visible, click Add to save the action
  7. Publish the action and add it to an entitlement
  8. Test by selecting a machine under Items and “Add VM to a Security Group” from the Actions menu

    VM Resource Action for Security Groups

    VM Resource Action for Security Groups

  9. You’ll be presented with the list of allNSX Security Groups to which you can add the selected VM

    Select Security Group

    Select Security Group

Conclusion

This part of the series should help streamline the management of VMs and their membership in Security Groups.  Obviously, items like removing a VM from a Security Group or even removing a Security Group are not included here.  The NSX plugin is missing quite a bit of functionality available in the API, so those additional functions require significantly more configuration.

Thanks to John Dias for his information and examples posted here.

 

Automating NSX Security Groups with vCAC/vRA – Part 1

In this series, I’ll document how to automate the creation and (some of) the management of NSX security groups within NSX.

First, what’s the use case?  Why is this interesting?  Let’s assume that you’ve decided to use large “flat” networks instead of many small networks.  One reason you may make that decision is because of the challenges with either having many blueprints (one per network!) or making changes to the workflows to reliably set the appropriate properties.

Background

In this solution, we’ll have to have vCAC 6.1 or vRealize Automation 6.2, NSX 6.x and vCenter/vRealize Orchestrator with the vCAC and NSX plugins installed and configured. We have two Logical Switches, one for Production and one for Non-Production.  In addition, there’s a corresponding network profile and the business groups have reservations.  Now, we have to ensure that there  are security boundaries within the flat networks.  We’ll accomplish this through Security Groups.

Caveats

We’ll create security groups and nod in the direction of security profiles, but will not be automating the creation of security profiles nor their assignment to the Security Group(s).  That can be done by the security admins through the NSX interface or maybe later we’ll add that capability too. 😉

 Procedure

  1. Create Security Groups.
    • Open vSphere Web Client and navigate to Networking and Security, then Service Composer.
    • Click the “New Security Group” icon
    • Enter a Name and Description for your new Security Group and click Next
    • If you want to create rules for dynamic membership or include/exclude existing VMs, you can do so in the subsequent steps.  Finish the wizard.
    • Repeat to create all of your security groups
    • Create Security Groups in NSX

      Create Security Groups in NSX

  2. Create Property Dictionaries invCAC/vRA.
    • Log into vCAC as an Infrastructure Admin and navigate to Infrastructure|Blueprints|Property Dictionary
    • Click “New Property Definition”, for the name enter “VCNS.SecurityGroup.Names.Production“.  You can replace “Production” with a name of your choosing, so you can have multiple lists.
    • Select “DropDownList” as the control type and check to make it required, click the green check to save.

      Create Property Dictionary

      Create Property Dictionary

    • Click the “Edit” link in the Property Attributes column
    • Click “New Property Attribute”, select “ValueList” as the attribute type
    • Set the name to something appropriate, such as the same name as the Property Definition or “ValueList” or “SecurityGroups”
    • In the Value field, enter the names of the security groups you want included.  Separate the group names by commas (no spaces).  If you have groups whose names include spaces or commas, put them in quotes.  Click the green check to save.
    • Repeat to create another property dictionary and attribute for the Non-Production list
  3. Update Blueprints.
    • Edit your “production” blueprints by adding the “VCNS.SecurityGroup.Names.Production” custom property. Set the value to your default security group or leave it blank to require a selection. Be sure to check the “Prompt User” box. Click the green check to save.

      Add Custom Property to Blueprint

      Add Custom Property to Blueprint

  4. Test
    • Submit a request for the affected blueprint and verify that the dropdown list of security groups looks like you expect it to. Remember, that unlike many other custom properties in vCAC (eg: Network Profiles), you CAN have multiple versions of this one and display different lists.

      Dropdown list of Security Groups

      Dropdown list of Security Groups

    • After a VMis provisioned, verify in the vSphere Web Client that ithas been assigned to the expected security group

      VM added to Security Group

      VM added to Security Group

Next

In the next parts of this series, I plan to address the problems of maintaining the dropdown list manually and having a single security group per machine.

Many thanks to my friend Grant Orchard for his article on selecting a security group in a blueprint . It was the inspiration for this series.

Quick and Dirty workaround for vCAC/vRA MSDTC issues

02/03/2015 Comments off

Please note that this is a workaround, not a fix. I couldn’t get the damn thing working reliably, so I bypassed authentication. I wish I knew why I got “DataBaseStatsService: ignoring exception: Error executing query usp_SeclectAgent Inner Exception: Error Executing query usp_SelectAgentCapablities” all the frigging time. Followed John’s advice and reinstalled MSDTC on e.v.e.r.y.t.h.i.n.g, no change. Made sure that the settings match VMKB2089583, no change.

As a workaround, I just selected “No Authentication required” on the Clustered MSDTC and the IaaS Manager Service services. Don’t do this as a long-term thing.

This works, but don't do it in production

This works, but don’t do it in production

Adding a vCloud Air Endpoint to vCloud Automation Center

08/21/2014 Comments off
vCAC Service Catalog

vCAC Service Catalog

So, you have vCAC configured locally and now have access to vCHS vCloud Air and want to create an endpoint to provision machines to your virtual data center.

In this example, the vCloud Air account has a Virtual Private Cloud.

Background

vCloud Hybrid Service was recently renamed to vCloud Air to reflect it changing services (and to get people to stop calling it vCHeeSe).  It uses a lot of vCloud Director terminology and behavior.

Preparation

  1. Make sure your vCloud Air account setup is completed.
  2. Login to your vCloud Air account
  3. Have credentials for an Infrastructure Admin on vCloud Automation Center

Get Information

What? Where? How?
vCloud Air Username and Password duh!  hello-my-name-is-nametag
Virtual Data Center Name On the Dashboard, under VIRTUAL DATA CENTERS, the top line in bold is the virtual data center name.

Virtual Data Center

Virtual Data Center

vCloud Director API URL Click the name of your virtual data center (it doesn’t give you the pointing-finger icon, but it is a link) to view the details.  On the right, under RELATED LINKS, click “vCloud Director API URL”.  Copy the value and paste it somewhere handy.

vCloud Director API URL

vCloud Director API URL

Network details  While still in the Virtual Data Center details, click the Networks tab to view a list of the defined networks.  By default you’ll have a “Default-Isolated” and a “Default-Routed”.  You can create more, but for this example, I’m just going to use the default-routed network. Locate the network and record the gateway IP.  Make a note of the network size (/24 in this case) and the IP Range.  The IP range is the range of values that vCloud Director/vCloud Air will assign to newly-created VMs in this network.

Default Routed Network Details

Default Routed Network Details

 

Configure vCAC

  1. Logon to vCAC Console as an Infrastructure Administrator
  2. Navigate to Infrastructure/Endpoints
  3. Click New Endpoint/Cloud/vApp (vCloud Director)
  4. In the Endpoint configuration, give it a catchy name like “vCloud Air” and description.
  5. In the Address field, paste the vCloud Director API URL, but only up to the :443.  Don’t need the path.
  6. In the Organization field, paste the Virtual Data Center Name – it’s case-sensitive.
  7. If you have Custom Properties to be applied, you can add them now or later.
  8. Click OK to save your endpoint

    vCAC Endpoint for vCloud Air

    vCAC Endpoint for vCloud Air

  9. On the list of Endpoints, mouseover the new one and choose Data Collection from the menu. On the Data Collection status page, click the Start button to begin data collection
  10. After the data collection completes successfully, you will be able to create or edit a fabric group to add the Virtual Data Center as a Compute resource.

Next Steps

  1. Create a network profile for the network(s) you recorded earlier.  I suggest configuring the vCAC Network Profile IP Range to be exclusive of the vCD IP range
  2. Create reservations for your business groups on the vCloud Air Compute Resource
  3. Consider a Reservation Policy for your vCloud Air reservations

Coming Soon

I’m hoping to configure the VPN connection to extend my lab network to the virtual data center. Creating blueprints for use on vCloud Air, making existing blueprints work on vCloud Air.

Thanks to David Hill for his post about this very topic

 

Configuring Replicated vPostgres for vCAC 6.x

08/14/2014 Comments off

This is the second in my series for building a fully distributed vCAC deployment.  In this part, we’re building the vPostgres database server with replication for use with vCAC 6.x.

I’m using v9.2.6.0. The vCAC 6.0 Support Matrix says 9.2.4 is supported but the PDF version of the Installation and Configuration guide says 9.2.4 or higher is supported.  I originally wanted to use 9.3.2.0 because the documentation includes replication, but I’m unsure whether it’s officially supported with vCAC 6.x yet.  We’ll still configure replication though 🙂  I’m going to front-end the vPostgres nodes with a vCNS Edge Gateway load balancer so that in the case of a failure, we don’t have to reconfigure the vCAC appliance database connection.  Updated documentation shows that for vCAC 6.0, vPostgres v9.2.4 is supported, v9.2.6 and v9.3.4 were untested.  For vCAC 6.1, versions 9.2.4, 9.2.6 and 9.3.4 are supported.
Prerequisites:

  • Reserve IP addresses for the appliances and the virtual IP.
  • Add DNS entries for the IP addresses.  I used vpostgres1 for vpostgres2 for the appliances and vpostgres as the virtual/load-balanced name/address.

vProgres setup Steps

  1. Download the VMware vFabric Postgres Appliance from my.vmware.com.
  2. Deploy the vFabric Postgres Appliance from OVF twice.  I named them vPostgres1 and vPostgres2.  vPostgres1 will be the master and vPostgres2 will be the slave.
  3. Power on vPostgres1, browse to https://vpostgres1:5480, logon as root using the password you entered during deployment.
  4. Configure the hostname (eg: vpostgres1.ragazzi.lab) and timezone
  5. Browse to https://vpostgres1:8443, leave the default values, enter your password and click “Connect” to enter the Enterprise Manager (vpgdbem)

    Login to vpgdbem

    Login to vpgdbem

  6. Click on localhost:5432/DB Login Users to list the existing users (just “postgres” so far)
  7. Click the green plus to add a new DB Login user.  In the properties, enter “vcac” (or whatever you want) as the name, check “Enable login”, do not check “Can create DB login users” and set the password.  Click OK to save.

    Create vcac user

    Create vcac user

  8. Click on localhost:5432 to display the overview and list of databases (just “postgres” so far)
  9. Click the green plus to Create a new database.  Just enter “vcacdb” (or similar) for the name, set the Owner to “vcac”, add a comment if you wish and click “OK” to save. Click the refresh button (blue ccw arrow) to refresh the list.

    Create vcacdb database

    Create vcacdb database

  10. Expand the Databases item under the treeview and select your new “vcacdb” database.  The database overview should load, displaying the uptime, size and more.

    Select the database

    Select the database

  11. Toward the right side of the window is a button labelled “Enter SQL”, click it.
  12. In the SQL Script area, type the following:

    CREATE EXTENSION "hstore";
    CREATE EXTENSION "uuid-ossp"

    SQL Statements

    SQL Statements

  13. Click “Execute” and check the Output|Messages area for SQL query succeeded
  14. Click the X to close the SQL window
  15. Complete steps 2-4 for vpostgres2. Do not configure any users or databases on vPostgres2.
  16. SSH into vpostgres1, logon as “postgres”, not root.
  17. Run this command to create a replication user named “replicate”:

    v9.2/opt/vmware/vpostgres/current/share/create_replication_user replicate

    v9.3/opt/vmware/vpostgres/current/scripts/create_replication_user replicate

    You’ll be prompted for a password and confirmation.

    Create "replicate" user

    Create “replicate” user

  18. SSH into on vpostgres2 as “postgres”
  19. Run this command to configure vpostgres2 as a replica:

    /opt/vmware/vpostgres/current/share/run_as_replica -h 192.168.101.31-b -W -U replicate

    Obviously, replace the red text with the IP address of the master vPostgres server. First you’ll be prompted for the password for the “replicate” user, then you’ll confirm the authenticity of the connection, then you’ll be prompted to enter the password for the postgres user on the master. Next, you’ll confirm that you want to enable WAL archiving on the primary/master by typing “yes” and lastly, you’ll confirm your intention to overwrite the data directory with the databases from the master. It’ll copy the tablespace over.

    Configure replica and confirm

    Configure replica and confirm

  20. Run this command on vpostgres1 to verify the replication:

    /opt/vmware/vpostgres/current/share/show_replication_status

    Replication Status

    Replication Status

Load-Balancer setup steps
I’m going to use the load-balancer feature in vCloud Networking and Security Edge Gateway. It’s not the most intelligent Load-Balancer ever, but it’s what I have.

  1. Using vCNS Manager, locate the appropriate edge gateway, click Actions|Manage to open it for editing
  2. On the Configure Tab, edit the interface that will listen on the virtual IP
  3. Edit the Subnet and add the Virtual IP. It’s probably not the primary IP. Save and publish those changes

    Add the virtual IP to the Edge Gateay

    Add the virtual IP to the Edge Gateay

  4. On the Load Balancer tab, on the Pools page, click “Enable”, then “Publish Changes”Enable Load Balancer
  5. Click the green plus to add a load-balancing pool
  6. Enter a recognizable Name and Description, click “Next”
  7. On the Services step, check only TCP, set Balancing Method to “ROUND_ROBIN” and the Port to 5432. Click “Next”
  8. On the Health Check step, set it as shown. Click “Next” when done.
  9. On the members step, click the green plus to add the IP address of you SSO servers to the pool. Add the primary/master vPostgress server with a weight of 1 or higher.  Add the slave/replica with a weight of 0 (zero).  This will ensure all of the traffic goes to the primary until it is changed in the event of a primary failure. Keep the TCP port and Monitor Port at 5432 for each. Click “Next” once all you members are added.
  10. Review the Ready to complete step and click “Finish” if it all correct
  11. Click the Publish Changes Button before proceeding
  12. Click the “Virtual Servers” link, then the green plus to add a Virtual Server
  13. Enter a meaningful name and description, provide the Virtual IP adddress that you added to the edge earlier, select the Pool created in the steps above and Enable TCP on port 5432. Make sure the “Enabled” box is checked. Click “Add” then “Publish Changes”

Now, when you configure your vCAC Appliance, provide the host name that resolves to the virtual IP address.vpostgres-vCAC1

Dealing with a failure
By default, the replica acts like a read-only copy of the database. It has a very short replication delay, so do not count on it to save you if you delete things from the primary.

When to promote a replica:

  • You’ve screwed up the network settings on the primary vPostgres node beyond repair; preventing vCAC from using it and replication from occurring
  • You’ve applied an update to the primary vPostgres node that broke it; preventing vCAC from using it and replication from occurring

When to NOT promote a replica

  • You deleted a bunch of stuff from vCAC. Too late! Those changes have already replicated
  • The physical host where the primary vPostgres virtual appliance was running has failed. Just wait for vSphere HA to being it back online
  • You want to see it run active/active. It does’t do that. relax

Recovery Procedure

  1. See if the primary/master node is up. If it is, stop here.
  2. Using the vCNS Manager web interface, edit the load-balancing pool, setting the weight for vpostgres1 (which has failed) to 0 (zero) and the weight for vpostgres2 (which we’re going to promote) to 1. Save and publish changes.
  3. SSH into the slave (vpostgres2) as postgres
  4. Run this command to promote the slave to master:

    /opt/vmware/vpostgres/current/share/promote_replica_to_primary

    The response will be “server promoting

  5. If/When vpostgres1 comes back to life, you’ll need to configure it as a replica to vpostgres2. Do this by running the command from step 19 above.
  6. Now if you want to make vpostgres1 primary again, I strongly suggest you stop the vcac_service on your vCAC appliances. Then, you’ll just promote it like you did before and make vpostgres2 a replica again.