ClusterLion

advertisement
ClusterLion
Robert Graf | CEO
Mobile +43 664 1314403
Email: rg@prolion.at
Overview
1. ProLion-NetApp Alliance
2. High Availability
3. Metro Cluster Challenges
4. ClusterLion Solution
5. ClusterLion vs. Common Solutions
6. ClusterLion Technology
7. Customer Benefits
8. References
NetApp Alliance
 ProLion CEO Robert Graf: former NetApp Country Manager
in Austria, 7 Years @ NetApp
 ClusterLion only offered for NetApp MetroCluster
 NetApp Alliance Partner
 EU Distribution Partner: Arrow ECS
High Availability in IT
 High Availability is a MUST in today’s IT world!
 This applies accross industries
 Mission –critical applications must be available at all times!
 Therefore, permanent IT availability “Always ON” is a
prerequisite for many companies and no longer an option.
 Any downtime costs money and image
Cost of Downtime
 The cost by industry and studies vary, but it is clear that IT
downtime causes considerable damage!
Split-Brain Syndrome
 Wikipedia: Split-brain indicates data or
availability inconsistencies originating from
the maintenance of two separate data sets
with overlap in scope, either because of
servers in a network design, or a failure
condition based on servers not
communicating and synchronizing their data
to each other.
 High-availability clusters usually use a
heartbeat private network connection which is
used to monitor the health and status of each
node in the cluster. For example the splitbrain syndrome may occur when all of the
private links go down simultaneously, but the
cluster nodes are still running, each one
believing they are the only one running.
The Challenge of Every Storage Cluster
 Every storage vendor on the market needs a quorum, witness
or tie-breaker to run automatic switchover in case of sitefailure!
 Expensive infrastructure investments in a 3rd data center
location and highly redundant interconnects form the primary
data centers to the quorum site are required!
 No infrastructure investment is needed, which offers the lowest
possible TCO for automatic switchover.
 ClusterLion is only available for NetApp MetroCluster.
7 Mode or cDOT 2-Pack MetroCluster
stretched HA
system01 failed !
cf giveback
takeover!
Srvc (b)
Srvc (a)
A/A Controller Failure Scenario
1.
2.
3.
4.
1st Controller fails
Identity „moves“ to 2nd controller
I/O passes through 2nd controller
After repairing1st controller,
issue „cf giveback“
5. Identity „moves“ back to 1st controller
6. Normal operations continue
7 Mode or cDOT 2-Pack MetroCluster
stretched HA
Srvc (a)
Srvc (b)
SiteA down or
cfcftakeover
-d
giveback
site-connection broken?
MC Site Failure Scenario
1. Entire Site A fails
2. 2nd controller checks heartbeat, diskconnections and IP connection while
still serving its data
3. Human or process on 3rd Site identifies
site-failure
4. Issue „cf takeover –d“
5. Identity „moves“ to second controller
cDOT 4-Pack MetroCluster / local HA
local HA
stretched HA
local HA
MC Fabric
Srvc(a)
Srvc(b)
NO AUTOMATIC SWITCHOVER BETWEEN DATA CENTERS
ONTAP 8.3 MetroCluster DR Guide
Source: http://mysupport.netapp.com/documentation/docweb/index.html?productID=62093&language=en-US
ClusterLion – The Solution
ClusterLion
“Switchback”
SRV1
SRV2
Customer Support
during Giveback
Switchover
Ethernet / SAN
2x RS232
2x Ethernet
Srvc(b)
2x RS232
Srvc(a)
2x Ethernet
Srvc(b)
MC Fabric
A2 A1
Grid
100m
B2 B1
UPS
Grid
Q
Open Ticket
Partner Helpdesk
Remote Quorum
100m
UPS
Use
Case: Power Outage
Monitoring:
1.
Reporting:
2.
Action:
Power
A2:
Active
Supply
Controller
Heartbeat
• B2:
Power
Off
Storage
A1:
Lost Controller
Cluster
• B1:
Power
Off Partner, NVRAM etc.
Partner
B2:
No Controller
Status
Heartbeat
• A2:
Active
Controller
Heartbeat
Heart-Beat
B1:
Controller
Error and Power Alarm
• A1:
Force Takeover
• Q: Open Helpdesk Ticket
MetroCluster Switchover
Available solutions for NetApp MetroCluster switchover:
TieBreaker
Manual Switchover
ClusterLion
Support for 7-Mode and cDOT
MC config.
✔
✔
✔
Continues operation even
during site-failure
✔
X
✔
Only two data centers are
needed to run switchover
X
✔
✔
Highly secure against SplitBrain and data loss
X
✔
✔
Independet remote view on
MetroCluster status
X
X
✔
Very easy to install and
operate
X
✔
✔
ClusterLion Technology
ClusterLion Technology
 ClusterLion without Front Cover
 „Hot Swap“ Battery
ClusterLion Technology




4x Power Input
4x Power Output
2x Cooling Fans
2x 24V Output for UMTS
Gateways
 Reset Button
 2x Serial Consol Port
 6x Ethernet Connectivity
ClusterLion Premium Support
Premium Support Contract:
 24x7 Phone Support
 Proactive notification of the Customer
 Automatic support ticket at Storage Vendor
 Support during cluster giveback
 European Maintenance Partner: Econocom
Osiatis
Customer Benefits
 ClusterLion increases the availability of your NetApp
MetroCluster.
 Even in the event of a total failure at one location, cluster
services are properly delivered. All applications remain available.
 ClusterLion works with only two locations. This reduces costs
and complexity.
 A third site (Quorum) will be provided by ProLion free of charge.
 ClusterLion prevents data corruption in case of Split Brain
syndrome.
 ClusterLion permanantly ensures a consistent state in the
storage cluster.
 ClusterLion can be retrofitted at any existing storage cluster.
The
not if to
you
...ifquestion
you canisafford
Thank
you!
can operate
afford
ClusterLion,
without
but...
ClusterLion.
Download