Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2001-08-27
2003-11-04
Coby, Frantz (Department: 2171)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000, C711S161000, C711S162000, C711S165000
Reexamination Certificate
active
06643671
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to a system and method for synchronizing a remote data copy and, more particularly, to a system and method for efficient snapshot synchronization of a data copy using at least one accumulation remote copy trio consistency group including a source volume, accumulated write commands, an established peer-to-peer remote copy (PPRC) volume pair, and data consistency models.
BACKGROUND ART
With increasingly large amounts of data being handled in data processing systems, storage systems, such as disk or tape storage systems, are being used to store data. Some organizations rely heavily on data and quick access to the data. Disasters caused by environmental conditions, user errors, or application errors may occur in which access to the data is lost for some period of time. Mirroring or copying data to a secondary storage system from a primary storage system is currently employed for recovery purposes to minimize the time in which access to data is lost due to such a disaster.
In that regard, peer-to-peer remote copy (PPRC) is a synchronous copy mechanism that creates a copy of data at a remote or secondary storage system. The copy at the secondary storage is kept current with the data located at the primary storage system. In other words, a copy of the data located at the secondary storage system is kept in synch with the data at the primary storage system as observed by the user of the data. Volume pairs are designated in which a volume in the primary storage system is paired with a volume in the secondary storage system. Data transfer occurs in pairs in which data is transferred from a volume in a primary storage system to a corresponding volume in a secondary storage system, which together may be referred to as an established PPRC pair.
With a PPRC system, a data copy made to the secondary, or “recovery,” storage system occurs synchronously from a host point of view with write operations to volumes in the primary storage system. When data is written to the primary storage system, the data written to a particular volume is also written to a corresponding volume in the secondary storage system using a path to the secondary storage system.
Effecting a data copy from a primary volume to a secondary volume in a PPRC system may include an internal snapshot copying mechanism for copying all of the data of a source storage volume to a primary storage volume in a PPRC pair, which data is then migrated to the corresponding secondary storage volume. The internal snapshot copying mechanism makes a copy of at least one pointer to the data of a source volume, and the primary volume in the PPRC pair uses the pointers to access the data. The internal snapshot copying mechanism makes a copy of the data of a source volume to a primary volume of a PPRC pair by using pointers to the data in the source volume and then storing the pointers in a map. By using pointers, the internal snapshot mechanism can quickly copy the data from the source volume without affecting the access of a host to the source volume. The primary volume then transfers the data to the corresponding secondary volume without any host access interruption to the source volume.
There is, however, a large time difference between snapshot copies and synchronizing remote copies. That is, if a user wants to migrate a point-in-time copy of data from a source volume to a secondary volume in a PPRC pair by snapshot copying the source volume to the primary volume in the PPRC pair, the entire source volume is sent to the secondary volume, an operation which can take a very long time.
Further, as the number of PPRC volume pairs that are established and attempt to move from a duplex pending to a duplex state increases, system resources become increasingly degraded as duplex pending pairs are added. Cache space, processor cycles, and data paths are consumed while duplex pending. A duplex pending pair is a pair of corresponding volume pairs in which the system is attempting to copy the primary storage volume to the secondary storage volume. A duplex state pair is a pair of corresponding volume pairs in which the data from the primary storage volume has been copied to the secondary storage volume. Moreover, individual primary volume performance may be additionally affected because the host has to compete with the synchronizing task for access to the source volume. The synchronizing task is the process of migrating the source storage volume to the secondary storage volume.
Thus, there is needed an improved system and method for synchronizing a data copy. In such a system and method, when a user wants to snapshot copy from a simplex source volume to a PPRC volume pair in order to migrate backup data to a secondary system, rather than sending the entire source volume to the secondary volume, only data indicated by accumulated write commands would be sent. In order to make the snapshot copying and the migration of the data efficient, a bitmap would be used to signify the accumulated write commands. Advantageously, only the data indicated by those write commands would be snapshot copied and migrated to the secondary.
Such a system and method would preferably employ a group of three storage volumes in this operation. The first would be the source volume of the snapshot copy, which would accumulate the write commands in a bitmap. The next would be the target volume of the snapshot copy, which would be a primary volume of a PPRC pair and would receive the results of the write commands and a copy of the bitmap having the accumulated write commands. The final volume would be on the secondary system and would be the secondary volume of the PPRC pair. The three volumes would together comprise an accumulation remote copy trio. The source volume could be specified along with the establishment of the primary target-secondary PPRC pair, or configured through an operations panel. Upon establishment of the PPRC pair, an internal snapshot copy would synchronize the source and primary target volumes. The primary target volume would begin synchronization with the secondary volume by sending over the entire volume. The source volume would establish a bitmap and begin accumulating write commands received from a host. Subsequent snapshot copies from the source volume to the primary target volume would only snapshot copy data indicated by the accumulated write commands. Only that data indicated by the accumulated write commands would then be migrated to the secondary volume.
Such a system and method would thereby allow a user to make a point-in-time copy of data and very efficiently migrate that copy to a secondary system without impacting the source volume. The target of the snapshot copy would be the primary of a PPRC pair that would transfer only the tracks specified in the bitmap to the secondary volume. In such a fashion, the PPRC pair would become duplex much more efficiently because only the specified tracks in the bitmap would be sent to the secondary volume. The bitmap could represent granularity at a record, track or cylinder level. Such a system and method would thereby remove host impact to the source volume while data is being migrated to the secondary volume. As a result, very little response time degradation would be seen by the host. Still further, the more efficient migration of data to the secondary volume would reduce the time it takes to synchronize the volumes, consume less system resources, and reduce the time interval between potential snapshot copies for migration purposes.
Such a system and method could also be employed in storage systems that comprise multiple source volumes, each associated with an established PPRC pair. In that regard, users increasingly have databases that span multiple source volumes, and would like to migrate data to a secondary storage system also having multiple volumes for disaster recovery purposes as described above. Such migration should be as quick as possible to facilitate smaller incremental backups. In such systems, however, data consistency becomes more importa
Eskenberry Robert P.
Hostetter David G.
Johnson Jennifer
Milillo Michael S.
West Christopher J.
Brooks & Kushman P.C.
Coby Frantz
Storage Technology Corporation
Thai Hanh
LandOfFree
System and method for synchronizing a data copy using an... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for synchronizing a data copy using an..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for synchronizing a data copy using an... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3175600