Virtualization Adapted Adapting Business Processes for Virtual Infrastrcuture (and vice-versa)

2010/04/20

vSphere Network Isolation Addresses

Filed under: virtualization — Tags: , , , , , , — iben @ 14:45

http://www.vmware.com/pdf/vsphere4/r40_u1/vsp_40_u1_availability.pdf

Network Isolation Addresses

A network isolation address is an IP address that is pinged to determine if a host is isolated from the network. This address is pinged only when a host has stopped receiving heartbeats from all other hosts in the cluster. If a host can ping its network isolation address, the host is not network isolated, and the other hosts in the cluster have failed. However, if the host cannot ping its isolation address, it is likely that the host has become isolated from the network and no failover action is taken.

By default, the network isolation address is the default gateway for the host. There is only one default gateway specified, regardless of how many service console networks have been defined, so you should use the das.isolationaddress[…] advanced attribute to add isolation addresses for additional networks. For example,  das.isolationAddress2 to add an isolation address for your second network, das.isolationAddress3 for the third, up to a maximum of das.isolationAddress9 for the ninth.

When you specify additional isolation address, VMware recommends that you increase the setting for the das.failuredetectiontime advanced attribute to 20000 milliseconds (20 seconds) or greater. A node that is isolated from the network needs time to release its virtual machine’s VMFS locks if the host isolation response is to fail over the virtual machines (not to leave them powered on.) This must happen before the other nodes declare the node as failed, so that they can power on the virtual machines, without getting an error that the virtual machines are still locked by the isolated node.

For more information on VMware HA advanced attributes, see “Customizing VMware HA Behavior,” on page 26.

das.isolationaddress
Sets the address to ping to determine if a host is isolated from the network. This address is pinged only when heartbeats are not received from any other host in the cluster. If not specified, the default gateway of the console network is used. This default gateway has to be a reliable address that is available, so that the host can determine if it is isolated from the network. You can specify multiple isolation addresses (up to 10) for the cluster: das.isolationaddressX, where X = 1-10. Typically you should specify one per service console. Specifying too many addresses makes isolation detection take too long and can affect VMware HA behavior.

das.usedefaultisolationaddress
By default, VMware HA uses the default gateway of the console network as an isolation address. This attribute specifies whether or not this default is used (true|false).

2009/03/24

Configure DataDomain NFS ESX DataStore

Filed under: virtualization — Tags: , , , , , , — iben @ 13:02

How to configure DataDomain as NFS DataStore for ESX
Purpose:
Follow these steps to allow a datadomain storage appliance to be used as a high capacity datastore by ESX and ESXi hosts using NFS so Virtual Machines can be moved from Production ESX cluster to test/dev ESXi hosts.

Basic steps:
1 – configure datadomain appliance nfs export
2 – add NFS datastore on ESX hosts
3 – copy VMs from Primary Storage to DDA (DataDomain Archiver).

Setup:
Data Domain OS 4.5.3.0-82657
Model 565
Active Directory Domain Admin user-id “itadmin”
DD IP Address: 10.10.10.99

Source ESX hosts’ VMkernel portgroups on 10.10.11.0/24 network
Destination ESXi host on 10.10.10.7 (VMkernel and Management are on shared Port Group)

itadmin@datadomain-lan# nfs show clients
path client options
—————- —————— —————————————-

itadmin@datadomain-lan# nfs add
nfs add [ ( ) ]
Add NFS clients to an export

itadmin@datadomain-lan# nfs add /share/vmtest 10.10.11.0/24
NFS export for “/share/vmtest” added.
dirodriguez@datadomain01-source# nfs add /share/vmtest 10.10.10.7
NFS export for “/share/vmtest” added.

itadmin@datadomain-lan# nfs show clients
path client options
—————- —————— —————————————-
/share/vmtest 10.10.11.0/24 (rw,no_root_squash,no_all_squash,secure)
/share/vmtest 10.10.10.7 (rw,no_root_squash,no_all_squash,secure)
—————- —————— —————————————-

Add share on ESX hosts.

Copy VMs from Source ESX SAN datastores to new shared datastore.

Copy VMs from shared datastore to local ESXi datastore and reconfigure.

Check Deduplication Compression.
You can see amount of deduplication provided by multiple VM images by running this command:

itadmin@datadomain-lan# file show compression /share/vmtest
Total files: 15; bytes/storage_used: 18.3
Original Bytes: 15,462,695,250
Globally Compressed: 2,191,220,844
Locally Compressed: 840,288,247
Meta-data: 6,953,000

In this example we are getting only 18x compression. With additional Virtual Machines the compression will increase allowing many VMs to be stored and archived.

Powered by WordPress