Electrical computers and digital processing systems: multicomput – Master/slave computer controlling – Master accessing slave storage
Reexamination Certificate
1999-01-06
2003-03-04
Sheikh, Ayaz (Department: 2155)
Electrical computers and digital processing systems: multicomput
Master/slave computer controlling
Master accessing slave storage
C707S793000, C707S793000, C707S793000, C707S793000, C711S159000, C711S162000
Reexamination Certificate
active
06529944
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to the field of mechanisms for allowing remote control to occur between host computers and a plurality of mass storage business continuance volumes and more particularly to cascading commands for issuance by a host system to collect information about and transmit control commands to volumes at one or more levels away from the host in the system.
2. Background
Mass storage systems have become increasingly cost effective for critical business systems. Advances such as Redundant Arrays of Independent Disks (RAID) technologies and Hierarchical Storage Management (HSM) systems have greatly improved the reliability of mass storage by providing a number of different redundancy features. Additionally, HSM systems such as the SYMMETRIX™ systems that are commercially available from the Assignee of the subject application provide disaster recovery facilities such as Symmetrix Remote Data Facilities (SFDF). These allow a SYMMETRIX™ 5xxx system located at one site to maintain a continuous copy or mirror of the data at the logical volume level in other SYMMETRIX™ systems located in physically separate sites.
FIG. 7
a
(Prior Art) illustrates a redundancy technique used in SYMMETRIX™ systems to provide mirroring, RAID configurations, and other forms of redundant disk storage. As seen in
FIG. 7
a
(Prior Art) disk adapters DA
1
, DA
2
and DA
3
are connected over small computer storage interface (SCSI) buses to physical disk drives C, such as C
1
, C
2
and C
3
on disk adapter DA
1
. In SYMMETRIX™ systems, in some implementations, a physical disk C
1
is divided into three logical disks, called H
1
, H
2
, and H
3
.
To illustrate this, assume a typical physical disk connected to a mainframe computer contains 2000 cylinders. In an HSM system such as SYMMETRIX™ systems, shown in
FIG. 7
a
(Prior Art) using disks C which are larger in capacity than the typical disks, the larger disks C can be logically divided into smaller logical units. If the disks C in this example hold 6000 cylinders, this physical disk C has the capacity of three typical disks. Each logical disk H, in this example, would be the equivalent of one typical disk.
Still in
FIG. 7
a
(Prior Art), if a typical disk contains a single large file or dataset named DSN
1
, mirroring redundancy techniques used in an HSM such as SYMMETRIX™ systems, can create copies of this dataset DSN
1
on disk adapters DA
2
and DA
3
. In the example here, the first standard copy of DSN
1
is physically located on disk adapter DA
1
, disk C
1
, at logical disk H
2
. The first mirror copy, DSN
1
M
1
is located physically on disk adapter DA
2
, at physical disk C
2
, logical disk H
1
. The second mirror copy DSN
1
M
2
is located on disk adapter DA
3
, physical disk C
3
, logical disk H
3
.
In
FIG. 7
b
(Prior Art), a more abstract way of thinking about mirroring or redundancy is shown. If the HSM system allows three mirrors for a standard disk, the HSM might have disks configured as shown here—the standard disk for data DSN
1
, is allocated to disk adapter DA
1
, physical disk C
1
. The first mirror, Mirror
1
, is assigned to disk adapter DA
2
, physical disk C
3
, and so on. In SYMMETRIX™ systems, the combination of disk adapter DA, physical disk C and logical disk H, is resolved into a SYMMETRIX™ device number. In this example, the SYMMETRIX™ system synchronizes the mirrors in a transparent manner. When the data from disk adapter DA
1
, physical disk C
1
has been copied to mirror
1
and mirror
2
, the devices are considered synchronized.
Now turning to
FIG. 7
c
(Prior Art), disk adapters DA are shown as they might be configured in Symmetrix Remote Data Facilities (SRDF) systems for disaster recovery. As seen in
FIG. 7C
(Prior Art), a SYMMETRIX A(MASTER) system has been configured in a unidirectional SRDF campus solution with SYMMETRIX B(SLAVE) system.
When the SRDF features are used, a SYMMETRIX™ system includes not only cache memory, channel directors CD and disk adapters DA, but remote link directors RLD. Within each SYMMETRIX™ unit, three volume types may be defined: local (L), source (R
1
) and target (R
2
). Local volumes L are volumes unique to that SYMMETRIX™ unit. They are only accessible to hosts attached to that SYMMETRIX™ unit (in this example, HOST
1
.)
Still in
FIG. 7
c
(Prior Art), source volumes R
1
are logical volumes that reside on a SYMMETRIX™ unit with the SRDF features activated, so that the data on source volumes R
1
is mirrored or copied to respective target volumes R
2
on another SYMMETRIX™ unit (in this example, HOST
2
). The target volumes R
2
are located on one or more separate SYMMETRIX™ units in an SRDF configuration.
As seen in
FIG. 7
c
(Prior Art), a path is established by the remote link directors RLD to allow data to be mirrored. The paths shown here are labeled remote access group
0
, or RA
0
.
Turning now to
FIG. 7
d
(Prior Art), a bidirectional SRDF configuration is shown. Host
1
is logically in communication with standard volume std, in this example. As mentioned above, SYMMETRIX™ systems would normally (in the absence of SRDF features) establish some mirroring for an ordinary volume. In this case, mirrors M
1
and M
2
in SYMMETRIX™ A might be established for standard volume std. The SRDF feature takes this mirroring one step further. Instead of creating a mirror of SRDF standard source volume R
1
on SYMMETRIX™ A, the SRDF features, using the remote link directors RLD, assign a remote mirror M
1
in SYMMETRIX™ B. In other words, source volumes R
1
are standard volumes that reside on a SYMMETRIX™ unit, with the data on those volumes mirrored to respective target R
2
volumes on another SYMMETRIX™ unit, here SYMMETRIX™ B. If the source volume R
1
fails, the SYMMETRIX™ A will transparently access the data on the corresponding target volume R
2
in SYMMETRIX™ B. When the failing volume is replaced, the remotely mirrored pair is re-synchronized automatically as a background operation, using the data in the target volume R
2
.
Still in
FIG. 7
d
(Prior Art), target R
2
volumes are a copy of the source R
1
volumes in another SYMMETRIX™ unit. A target volume R
2
, typically has a default configuration mode of “read-only” to any host with access to the SYMMETRIX™ unit in which it resides. In this example, mirror M
1
, in SYMMETRIX™ B, which is the remote mirror to the source volume R
1
in SYMMETRIX™ A, would be made “not-ready'” to Host
2
. Normally, writes to the target volume R
2
on Host
2
, occur via the link paths created by the remote link director. However, if the source volume R
1
on SYMMETRIX™ A fails, SYMMETRIX™ A will transparently access the data on its corresponding target R
2
volume (here, M
1
in SYMMETRIX™ B)
A target volume R
2
typically has a default configuration mode of “read-only” to any host with access to its SYMMETRIX™ unit, in this example, HOST
2
. To enable disaster recovery, the remote link directors RLD in a SYMMETRIX™ unit, create link paths RA
0
with the SYMMETRIX™ unit containing the target volumes R
2
. As data is written to a source volume R
1
by HOST
1
, the remote link director RLD in SYMMETRIX™ A automatically writes a copy of that data over link path RA
0
of the SRDF connection to the corresponding target volume R
2
on the designated SYMMETRIX B system. Thus, if a source volume R
1
fails for any reason, the remote link director RLD will transparently access the data on the corresponding target volume R
2
on the designated target SYMMETRIX B system over the link paths RA
0
and transmit that to HOST
1
.
New writes to the failed source volume R
1
in SYMMETRIX™ A accumulate as invalid tracks in the cache of SYMMETRIX™ A—the unit containing the source volume R
1
. When the failing source volume R
1
is replaced, the remotely mirrored pair is re-synchronized automatically as a background operation within the SYMMETRIX™ unit, using the data in the appropriate target volume R
2
.
FIG. 8
(Prior Art) shows an extended distance implementation of the SRD
EMC Corporation
Gunther John M.
Jean Frantz B.
Sheikh Ayaz
Wilson Penelope S.
LandOfFree
Host system for remote control of mass storage volumes using... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Host system for remote control of mass storage volumes using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Host system for remote control of mass storage volumes using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3058827