You are here: Home > Solutions > IPSTOR ® Disaster Recovery

IPSTOR ® Disaster Recovery
Protect from what could never happen... But did!

In today's corporate 24x7xForever uptime environment, a practical Disaster Recovery (DR) solution must fulfill the following objectives:
  • Address all potential disasters including sudden outages and rolling disasters 
  • Achieve established recovery time objectives (RTO) and recovery point objectives (RPO)
  • Provide effective data protection during each phase of DR:
    pre-, during, and post-disaster 

Power outages, fire, flood, etc., can cause sudden unplanned outages that wreak havoc on day-to-day operations. However, a rolling disaster, which is defined as “gradual data corruption that happens over a prolonged period of time, undetected until a catastrophic failure finally occurs,” is particularly difficult to recover from, because by the time it is detected, it can be too late to recover – the data at the DR site could already be corrupted. 

Key Benefits
  • Delivers rapid resumption with minimal data loss
  • Addresses both sudden outages and rolling disasters
  • Offers host/OS and storage hardware independence


There is a solution: The IPStor ® Disaster Recovery Software Solution, consisting of Synchronous Mirroring; Asynchronous Mirroring; and database-aware, transaction-based Replication, defends against a rolling disaster without the expense or integration of additional software or hardware.

And, whether it's to achieve established recovery point objectives (RPO) or recovery time objectives (RTO), the IPStor ® DR Solution provides rapid resumption of business in the event of a sudden unplanned outage with minimal data loss, by delivering the functionality required to comprehensively address all phases of a disaster: pre-, during, and post disaster, as follows:

Phase 1: Pre-Disaster


LUN Pre-assignment Prevents Chaos When a Disaster Strikes – IPStor ® software allows a replica disk to be pre-assigned to hot standby machines at the DR site, but prevents them from writing to the replica. If a disaster strikes, the replica is promoted to primary status, and the LUN assignment operates automatically. This prevents the chaos usually caused when Administrators are not sure which replica should be assigned to which standby host during an emergency.

Fast Initial Synchronization – In the case where the primary disk is very large (e.g. 5TB), it would take more than 31 hours to replicate 5TB over a DS3 (45Mb/s) line. To solve this dilemma, IPStor ® software allows bulk-transfer of the replica via either mirror-and-ship, or by image backup/restore via tape, so that the replica is in sync except for changes that may have occurred during the shipping or backup period. Then IPStor ® uses a ‘scan-for-differences' method to generate the initial delta-map, and rapidly syncs up the small amount of difference. 

Provides data copy sets within or across cabinets, delivering comprehensive fault tolerance

Phase 2: During Disaster


Sudden Disaster – If the primary site suddenly fails, it could be in the middle of a replication process. IPStor ® defends against replica corruption by storing the changes in a separate area, and does NOT write the changes to the replica unless the entire data set is received and confirmed. Therefore, either a replication is successful or it has failed and will be retried. No partial changes are applied to the replica. 

Rolling Disaster – If the primary site is gradually failing, this may lead to corruption that is not detected until it is too late. But when the IPStor ® point-in-time (PiT) imaging option ( TimeMark ) is enabled at the replica, each individual change is journaled. When the disaster is finally recognized, the replica can be promoted, and the TimeMark PiT snapshots can be used to go back in time until a good state is found. IPStor ® can keep 256 PiT images, which can represent weeks of historical data for rollback purposes. Most legacy replication products (i.e. Veritas Volume or File Replicator, NSI Double-take, etc.) do NOT have the ability to keep track of historical replication data, and can therefore be vulnerable to a rolling disaster. 

Delta based transaction-aware replication that enables sites to replicate data across any distance and delivers rapid recovery in the event of a sudden or rolling disaster


Minimum Data Loss – The IPStor ® Asynchronous Mirroring option can be used as part of a DR scheme to provide a low-loss mirror copy of the primary. Although this copy might take longer to start up (requiring Checkdisk, FSCK, database recovery, etc.), the advantage is that it is nearly in sync with the primary, so very few (if any) transactions are lost. 

During a disaster, the replica is used for instant restart (enabling business to continue), while the mirror copy completes the recovery process. When completed, the mirror can be used to recover additional transactions not found in the replica.

Provides remote data copy sets within or across cabinets for environments consisting of two systems located at geographically remote sites


Rapid Restart – FalconStor also offers IPStor ® Snapshot Agents for databases and message stores (Oracle, SQL, Sybase, DB2, Microsoft Exchange, Lotus Notes/Domino...) to ensure that the replica is always in a consistent state and ready to be promoted and mounted by the OS and the database/messaging application. Without these IPStor ® Snapshot Agents, there is no way to guarantee the replica's data integrity, and it will be necessary to perform time-consuming Checkdisk, FSCK, and DB recovery procedures before re-starting the application.

Emergency Access over IP – If the primary site loses only the storage, but the servers are still healthy, it is possible to use IPStor ® to access the promoted replica at the DR site.

Phase 3: After Disaster



Fast Reverse-Sync – To revive the primary site after the problem is fixed, it is necessary to reverse synchronize newer data from the DR site back to the primary site. IPStor ® facilitates role reversal and “scan-for-difference” operations to minimize the time necessary to re-establish the primary site.


IPSTOR ® DISASTER RECOVERY SOLUTIONS:

IPStor ® Delta-Based Replication Highlights
Point-in-time copy for rapid data recovery

  • Zero performance impact on the host(s) at primary site
  • Improves performance and utilization of existing communications link via flexible policy to trigger replication
  • Database transactional integrity and versioning maintained at secondary site
    • Minimal failover time
    • Up to 256 versions maintained per Replica
    • Current/prior version of Replica can be mounted for Read/Write access over IP/FC
  • Applies “changed blocks” to the database at the primary site
    • Minimal fail-back time
  • Data compression to maximize network bandwidth
  • Data encryption for end-to-end security
  • Automatically recovers from network failure without restarting the replication session
IPStor Synchronous Mirroring Highlights 
Real-time high availability of data
  • Primary and online copy of data are identical and concurrent
  • Provides redundancy and automatic failover at disk, LUN, channel and cabinet level
    • Redundant copy of data is always current
  • No CPU impact to application servers
    • Application servers do not need to process two write commands
  • Uses mirror-and-swap technique
    • Eliminates downtime during disk upgrade and/or replacement
    • Enables efficient data migration from old to new storage
    • Enables efficient data migration from old to new storage
  • Fast and easy point-and-click mirror creation and definition via IPStor Console
  • Maintains access during initial mirror creation
IPStor Asynchronous Mirroring Highlights
Near real-time availability of data
  • Online standby of data (mirror) at secondary location is a few seconds behind primary data set (based on overall geographic distance
  • Provides redundancy and automatic failover at disk, LUN, channel, and cabinet level
  • No CPU impact to application servers
    • Application servers do not need to process two write commands
  • Provides benefits of mirroring with no performance impact at the primary site
    • Enables writes to reach mirror over long distances between data centers
  • Fast and easy point-and-click mirror creation and definition via IPStor Console
  • Maintains constant access to applications during initial mirror creation

Resources

Where to go from here?

Tell us about your project / your needs
Call us @ 1-800-963-6329 and ask to speak with a IPStor Disaster Recovery consultant.
Call Me Now
Chat online

 

Mendax Microsystems Inc. licenses the IPStor ® product. IPStor ® is a registered trademark of FalconStor Software Inc.






[Company]   [Contact Us]   [Support]   [Warranty]   [Discount] [Price Drop]  
[Credit Application]   [Sitemap] [Privacy Policy]   [Terms of use]   [Security]  

Copyright Mendax Microsystems Inc. 1996-2007 all rights reserved
USA including APO / AFO, CANADA & OVERSEAS
Toll Free: 1-800-963-6329
Back to Mendax Store

Search Engine Optimization by MetaSEO.com