The document discusses improving application availability on virtual machines. It notes that while server virtualization improves hardware availability, business applications still suffer downtime. It argues that application awareness is needed to prevent downtime. The document then describes how the software solution Neverfail adds application awareness and intelligent automation to protect applications from downtime. It also explains how Neverfail can be combined with virtualization tools like VMware SRM to maximize availability benefits.
Unleash Your Potential - Namagunga Girls Coding Club
Improve App Availability on VMs
1. Improving
applica.on
availability
on
virtual
machines
Abstract:
Server
virtualiza1on
enables
a
wide
range
of
high
availability
and
disaster
recovery
op1ons,
most
of
which
focus
solely
on
the
virtual
hardware
and
leave
applica1ons
vulnerable
to
down1me.
– In
this
thirty-‐minute
discussion
we’ll
cover
such
points
as:
– Why
server
virtualiza1on
works
for
hardware
availability
– Why
business
applica1ons
s1ll
suffer
down1me
– How
Neverfail
protects
applica1ons
from
down1me
– How
to
combine
Neverfail
SRMXtender
&
VMware
SRM
1
3. Agenda
– Why
server
virtualiza1on
works
for
hardware
availability
– Why
business
applica1ons
s1ll
suffer
down1me
– The
role
of
applica1on-‐awareness
in
preven1ng
down1me
– How
Neverfail
adds
incremental
value
to
virtualiza1on
tools
– Where
to
learn
more…
3
4. What
should
be
the
scope
of
an
availability
strategy?
– Keeping
the
power
on?
– Keeping
the
data
protected?
– Keeping
the
computer
up
and
running?
– Line-‐of-‐business
applica1ons
have
one
simple
need:
Minimizing application downtime
and keeping users connected
to their applications and their data
regardless of the mode of failure
4
5. Yet
the
most
likely
failure
modes
are
least
protected
Source: Freeform Dynamics
5
6. Virtualiza.on
has
enabled
new
availability
strategies
– The
“machine”
is
now
abstracted
from
the
“hardware”
– Makes
the
machine
more
“portable”
– Image
files
can
be
copied,
shared,
and
replicated
– Enables
faster
backup
&
restore
of
VMs
– VM
images
can
run
on
any
hypervisor
host
– Doesn’t
maZer
which
hardware
vendor
you
choose
– Hypervisors
typically
need
to
be
homogeneous
– Impact
of
physical
hardware
failures
can
now
be
minimized
– Infrastructure
under
the
VM
image
has
become
more
resilient
– Computers
can
stay
running
longer,
but
yet
we
s1ll
have
down1me
6
7. Virtualiza.on
tools
focus
on
infrastructure
availability
– Backup
and
recovery
of
image
backups
– Restore
a
VM
image
to
another
host
and
restart
– Shared
storage
and
VMo1on
– Live
migra1on
of
a
running
VM
onto
an
adjacent
host
connected
to
same
storage
(enables
planned
maintenance,
load-‐balancing)
– Shared
storage
and
VMware
HA/FT
– Restart/sustain
a
VM
on
an
adjacent
host
connected
to
same
storage
(following
primary
host
or
opera1ng
system
failure)
– Replicated
storage
and
Site
Recovery
Manager
(SRM)
– Restart
all
or
a
group
of
VM
images
an
remote
datacenter
(following
a
site
disaster)
7
8. What
part
of
the
solu.on
stack
focuses
on
the
app?
Applica1on
Database
Server
File
Server
Mail
Server
Mobile
Device
Microsoc
Web
Server
Server
Server
SharePoint
Server
Applica1on
Availability
None of these solutions offer
Hardware
Availability
application availability
SAN
Physical
Hypervisors
SAN
Physical
Hypervisors
WAN
SAN
Storage
SAN
Storage
Primary
Site
Recovery
Site
VMo1on
/
SRM
/
Storage
VMware
HA
Replica1on
8
9. How
does
VMware
protects
its
own
cri.cal
app?
– VMware
vCenter
Server
Heartbeat
TM
– Ensures
Availability
for
the
VMware
vCenter
Server
Management
Pladorm
– “VMware
vCenter
Server
Heartbeat
delivers
high
availability
for
VMware
vCenter
Server,
protec1ng
the
virtual
and
cloud
infrastructure
from
applica1on,
configura1on,
opera1ng
system,
network
and
hardware-‐related
problems.”
– hZp://www.vmware.com/products/vcenter-‐server-‐heartbeat/
– (Did
I
men1on
that
Heartbeat
is
a
Neverfail
product?)
– Download
the
vCenter
Server
Heartbeat
Datasheet
9
11. Neverfail
improves
applica.on
availability
– Goes
beyond
moving
a
VM
image
from
one
host
to
another
– Not
propaga1ng
troubles
within
an
image
to
the
recovery
server
– Maintains
an
up-‐to-‐date
warm
stand-‐by
server
– Replicates
data
and
app
changes
from
within
the
machine
– Adds
applica1on
awareness
and
intelligent
automa1on
– Monitors
for
failures
across
the
en1re
applica1on
ver1cal
– Layers
incremental
value
on
top
of
VMware
SRM
– Any
physical
server
can
be
fully
integrated
within
SRM
Recovery
Plans
11
12. App-‐Awareness
Remediates
Most
Probable
Failures
Neverfail application awareness adds
detection of these failure modes
Replication vendors, VMware HA/FT, detect these issues
Source: Freeform Dynamics
12
13. Combining
Solu.ons
to
Maximize
Benefits
– Hypervisor
provides
local
hardware
availability
– Neverfail
provides:
– Applica1on-‐aware
protec1on
and
recovery
orchestra1on
– Support
for
physical
servers
and
different
hypervisors
within
SRM
– Bandwidth
op1miza1on
with
WANSmart
data
deduplica1on
WANSmart™
SAN
Hypervisor
Hypervisor
Hypervisor
13
14. Common
Concerns
Faced
By
SRM
Customers
– “I
want
to
standardize
on
SRM
to
reduce
costs
but
it
doesn’t
support
my
physical
servers”
– SRMXtender
integrates
failover/failback
of
physical
servers
within
SRM
Recovery
Plans
– Management
console
(VA
Director)
“plugs
into”
the
vSphere
Management
Client
– Customers
without
SRM
can
also
use
tool
to
protect
their
servers
– “I
want
to
protect
my
cri1cal
apps
with
low
RTO/RPO
but
SRM
can’t
give
me
the
recovery
performance
I
need”
– Neverfail
failover
technology
fails
over
with
<2min
RTO,
<5
sec
RPO
– SRM
offers
RTO
>30
mins
– SRMXtender
plugs
fast
failover
into
SRM
Recovery
Plans,
Creates
“Priority
Lane”
for
cri1cal
apps
– SRMXtender
enables
fast
failover
for
cri1cal
apps
standalone
as
well
14
15. SRM
Recovery
Plans
can
now
include
physical
servers
Corporate Messaging System
Web Front End
VMware Site Recovery Exchange Exchange Exchange
Virtual
Manager CAS CAS OWA
Application Tier
Inject
recovery
step
for
physical
servers
Exchange BES Anti-virus Anti-spam
Hub Server Server Server
Database / Storage Tier
Physical
SRMXtender MBX Server SQL Server File Server
15
16. Seamless
UI
integra.on
with
VMware
vCenter
Server
1.
Integrated
with
latest
vCenter
5.0
2.
Supports
physical
servers
3.
Supports
business
applica1ons
4.
Integrated
with
VMware
Site
Recovery
(SRM)
16
17. SRM
coordinates
physical
server
failover
sequence
SharePoint Farm SharePoint Farm
Web Tier Web Tier
Virtualized
IIS+MOSS IIS+MOSS IIS+MOSS IIS+MOSS IIS+MOSS IIS+MOSS
Array Replication or
Application Tier Application Tier
vSphere Replication
Virtualized
Query Index Excel Calc Central Query Index Excel Calc Central
Server Server Services Admin Server Server Services Admin
Database Tier Neverfail Replication Database Tier
Physical
& Failover/Failback
SQL Server SQL Server SQL Server SQL Server
Primary Site Secondary Site
Co-‐ordina7on
VMware
Site
Recovery
Manager
17
18. Key
Take-‐Away’s
– Virtualiza1on
provides
excellent
hardware
availability
– However,
user-‐error
and
applica1on
failures
are
the
most
probable
cause
of
down1me
– Neverfail’s
applica7on-‐awareness
detects
&
remediates
the
most
probable
failure
modes
– Neverfail
adds
incremental
value
to
vSphere
via
integra1on
with
the
vSphere
client
and
VMware
Site
Recovery
Manager
18
19. Visit
us
Online
to
Learn
More
– See
SRMXtender
&
the
Virtual
Availability
Director
– Video,
datasheets,
and
download
registra1on
– hZp://srmx.neverfailgroup.com
19