MATH 311: COMPLEX ANALYSIS |PREVIEW LECTURE

By
Published by

  • cours magistral
  • cours magistral - matière potentielle : d1
  • expression écrite
MATH 311: COMPLEX ANALYSIS — PREVIEW LECTURE 1. Introduction Complex analysis: not very complicated, not much analysis. The analysis is what's called “soft” analysis – some integrals and derivatives, but very few epsilons and grungy estimates once Cauchy's Theorem is proved. Real Analysis (Math 321) is not a prerequisite for this course, but students who haven't had it will sometimes have to accept results and methods on faith or do extra background work.
  • models for hyperbolic geometry
  • rational function
  • conformal map ω
  • riemann
  • many primes
  • area of a right triangle with rational sides
  • integrals
  • region
  • function
Published : Tuesday, March 27, 2012
Reading/s : 10
Tags :
Origin : sp.ts.fujitsu.com
Number of pages: 14
See more See less



Issue November 2011 BS2000/OSD Scope

Chapter 5
Use External System management

Pages 13

Introduction
IT running costs and IT availability are determined to a considerable degree by the quality of the system administration
functions of the used platforms.
The data center features of BS2000/OSD mainframes have been optimized over decades, as reflected by the benefits of
operational efficiency and low total cost of ownership.
In heterogeneous IT landscapes, however, individual platforms are no longer independent of each other. Many IT
resources, particularly disks, tapes and printers, are shared between several servers. Consequently it is extremely
important that system management functions can be integrated across Dynamic Infrastructures.
















Contents

Introduction 1
5.1 Overview 2
5.2 Concepts 2
5.2.1 Enterprise job scheduling 2
5.2.2 Enterprise print services 4
5.2.3 Data backup and restore 6
5.2.4 Performance monitoring and trend analysis 7
5.2.5 Management of intelligent disk subsystems 9
5.2.6 System resource accounting for the data center with BS2000/OSD 12
SNMP management in BS2000/OSD 12 5.2.7


Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 2 / 14
5.1 Overview
In the general context of system management in BS2000/OSD being considered here, suitable products are provided to assure
optimum manageability of the associated functionality. These cover 24-hour system monitoring and rule-driven responses, for
example to initiate fast application failover in the event of an error, backup of data resources using mirroring, and dynamic
system update options.
Ease of handling based on extensive automation and intuitive user interfaces are key factors contributing to the dominating
best-in-class position that BS2000/OSD occupies in respect of total cost of ownership, with running costs driven so low as to be
virtually unbeatable by rival systems.




Figure 1: BS2000/OSD System Management


The above figure shows system management classified into task areas and the main products assigned to these areas. A
priority in BS2000/OSD is of course to develop the associated products further according to their relevance to the manageability
of the overall BS2000/OSD system. The product portfolio has achieved a high standard of functionality and quality, and is being
constantly expanded to meet the new requirements arising from areas such as Dynamic Infrastructures and mobile connectivity.
The general aim is to improve cost-effectiveness and increase availability for the same or a reduced cost. This process is
supported by leveraging the advantages of the latest technological innovations.



5.2 Concepts
5.2.1 Enterprise job scheduling
Scheduled IT production processes that require no further dialog input once the current data has been assigned are referred to
as batch production. In real-world operation, many batch tasks need to be handled each day, and usually these exhibit a
complex interdependent structure. If the individual steps in batch production – the jobs – are handled manually, frequent job
aborts due to human error cannot be ruled out. While it is true that the individual steps and the necessary consideration of
dependencies can be managed by job chaining using customized procedures, the need to maintain and adapt the procedures to
new technologies as well as provide additional functions can often cause problems. For this reason, job scheduling products are
deployed today in order to facilitate efficient, system-based job processing. The strategic standard product for automatic job
control in heterogeneous configurations with BS2000/OSD is AVAS (job handling and processing system).

Transparency and fault tolerance are important basic principles of AVAS. All the parameters for controlling batch production are
entered not in the jobs themselves, but at the next higher level of the job nets – jobs and procedures can run without
modification.
Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 3 / 14
BS2000/OSD
HighlightsBS2000/OSD
 Scheduling of AVAS network
jobs on
Job010 Job011 BS2000/OSD,
Job012
Job020 Job021 UNIX-based
Job030 Job032 systems,
Windows, Linux, Job040
SAP R/3
Solaris (SPARC)  Centralized
Div. UNIX-based monitoring
systems *  Company-wide
dependency AVASSAP R/3 management
Windows
Linux
* = HP UX, AIX
other UNIX-based systems on request

Figure 2: AVAS


AVAS benefits
AVAS offers the following benefits:
Jobs are supplied with current data automatically, which means no risk of human error. Job nets are automatically started at
predetermined times. AVAS automatically initiates restarts if unexpected events occur. Manual activities can be largely
eliminated. Staff productivity increases.
AVAS enhances the self-optimizing characteristics of BS2000/OSD by increasing the efficiency of resource usage and reducing
the need for manual intervention.

 Jobs are started at precisely the time all dependencies are resolved. This also includes the complex control of job
parallelization. False starts as a result of unresolved dependencies are eliminated. It has been shown in practice that the job
error rate can be reduced from 10% to less than 1%. All batch processing can be automated to a large degree.
 Steadily increasing production with thousands of nets and jobs can be kept under control by combining groups of nets and
jobs into hypernets. Planning, production preparation and the release of subnets take place via the hypernet. The hypernet
concept makes it easy to specify the production sequence and provides a clear overview for production monitoring because
only the larger units of the hypernets need to be observed.
 Openness toward other platforms (Solaris SPARC™, Windows, Linux, HP-UX, AIX) means that cross-platform handling
processes can be monitored and controlled using a single tool. Specialist knowledge of each platform is not required.
In the cross-platform computer network, an exceptional level of automation is achieved in that jobs that are to be started on a
different, unavailable computer are first placed in a wait state and started automatically when the computer becomes
available again.
 Batch runs under the control of SAP’s R/3 enterprise resource planning software can also be integrated into the cross-
platform handling process. A SAP R/3 job is a set of batch run instructions defined using R/3 language elements that is saved
internally in R/3 and processed by R/3. The result of the run is then made available for further analysis.
 The connection to decentralized BS2000/OSD systems is implemented via TCP/IP by means of the product AVAS-SV for
BS2000/OSD (software selectable unit AVAS-SV-BS2): Starting and monitoring of a server job is also available for a remote
BS2000. A server AVAS-SV-BS2 that receives and executes jobs of an AVAS of another BS2000 computer runs in the
remote BS2000. AVAS-SV-BS2 monitors the job run, reports its termination to AVAS and signals and transfers the
corresponding job protocols to the AVAS database.
The BS2000/OSD systems involved can so be far more remote compared to the limits of the HIPLEX MSCF network.
 AVAS provides an extremely high standard of security by implementing admission and access control at different levels and
by audit-proof logging of all actions and interventions. Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 4 / 14
 All input data, events (normal/abnormal terminations, etc.) and interventions are logged without interruption. Reports are
made available for reporting and auditing purposes. To ensure effective, error-free development of job nets, various
interrelationships can be represented. A history file with compressed data relating to completed jobs and nets enables
analyses for statistical purposes and for planning support.

New functions of AVAS V8.5 (Current version)
 New calendar processing functions
The type NWRK (non-production day) has been extended to include WKND (weekend) and HLDY (holiday).
CREATE-CALENDAR defaults the missing entries from FIRST-/LAST-CALENDAR-DATE.
CREATE-/MODIFY-CALENDAR is used to specify the entries for SPECIAL NWRK OR FREE DATES with WILDCARD
syntax periodically for all the calendar years.
The PRINT function is implemented for SHOW-CALENDAR.
 ISAM protocol of journal file
All journal entries can be logged in an ISAM file. This ISAM file is opened SHAREUPD and can be used, for example, to
transfer journal entries to a database and process them using evaluation tools.
 Large log files can be transferred more effectively from WINDOWS/UNIX/LINUX servers to the AVAS-BS2000 system when
the transfer is via openFT in compressed form.
 AVAS supports the specification of IPv6 addresses when coupling server processors. In other words, IPv6 addresses can be
specified for the IP address of the server processor as well as the IP address of the processor(s) on which the server
monitoring and server interface processes run.

Outlook
New functions of AVAS V8.6 (planned release 4.Q 2012)
The following functional extensions are planned for the future version AVAS V8.6:
 Extension of the system limits:
The maximum number of plannable networks and networks that can be processed by AVAS-QUER is to be increased.
If it makes sense, optional parameters are provided as generally available options.
 New functions or extensions to existing functions, e.g.:
Entries in ABLDAT are made readable for databases.
In CREATE-PLAN-NET it is possible to select jobs with ADD.
InATE-ORDER support is provided for the parameter ALTERN-NET-NAME.
Improvement in the readability of the AVAS calendar
Introduction of the UNICODE mode in EDT operating mode.
 Revision of the AVAS-SV installation scripts for Linux/Unix.


5.2.2 Enterprise print services
It is still true that every business, in any sector, relies on printing to produce business-critical output, such as business letters or
invoices, technical descriptions or other documents. Since the advent of the paperless office was first announced, average print
volumes have increased rather than the reverse, while the trend toward distributed printing on heterogeneous enterprise
platforms with decentralized LAN workplace printers (host-to-LAN printing) has also strengthened. At the same time this has not
undermined the importance of centralized high-performance printing systems with their comprehensive paper post-processing
functions. Conversely, in fact, these are intensively being used by the users at the decentralized workplaces as well (LAN-to-
host printing).
The resulting complex diversity with its many different user and administration interfaces means a proportionally high
administrative overhead as well as huge potential for errors, with negative consequences for overall productivity if no
countermeasures are taken. Uniform, centralized print and output management is therefore a key requirement if a company
wants to make the best use of its print resources. An added benefit with such a solution is that security, so essential anyway for
business-critical output, is also automatically extended to all other (client) applications. In this context it is usually worth
reengineering the traditional print applications completely, integrating document management across the entire workflow to
produce a fully automated production process. This kind of holistic solution is also advocated today by all hardware
manufacturers and requires the perfect harmonization of all the hardware and software components involved. In view of the
permanent pressure to rationalize in IT departments, the greatest importance is also attached here to the requirement for use of
the extended output processing options without need for intervention in organically evolved legacy applications.

BS2000/OSD Spool & Print Services are ideally suited to meeting these requirements. Our partnership with Océ, one of the
leading manufacturers of print systems, means that we can provide complete hardware and software solutions which also take
account of current hardware trends such as multifunctional devices with color capability. The Spool & Print Services include the
local SPOOL of BS2000-GA and the following products:



Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 5 / 14
LANClient-ServerClient-Server
printer
UNIX based BS2000/OSD System TCP/IPMS WIN
EMDS
TDS
P RSO rdS SEH 3 Party Sys.
lpdOP (e.g. z/OS)
O
lpdLO
L D Wprint- lpd- pG
G rAA
i
ISO-DPAn Xprint
Router
t
PRISMAproductionServer
0 Copyright 2009 FUJITSU TECHNOLOGY SOLUTIONS
Figure 3: Spool & Print Services


Dprint
Dprint (Distributed Print Services) is the key product for implementing an enterprise-wide printing policy with BS2000/OSD as
the central print server in a heterogeneous network including Unix-based and Windows platforms. LAN-to-host printing by PC
clients in particular is perfectly integrated in the Windows interface, so that a centralized BS2000/OSD printer appears like an
additional powerful desktop printer. On the BS2000/OSD server side, the freely scalable high-performance print systems of
Fujitsu partner OPS (Océ Printing Systems GmbH) are used with the AFP/IPDS mainframe standard for production printing with
maximum reliability and performance requirements. Whether it comes sooner or later, the transition from traditional line printing
(E mode) to modern page printing (I mode) is unavoidable, even simply because of increasing quality demands (including higher
resolution, pixel-precise addressing). However, thanks to the comprehensive migration support available, this should no longer
pose any major problems. This migration should be completed before end of maintenance is reached for the old hardware.

RSO
The opposite route, i.e. printing from the host into the network to LAN or desktop printers, can be implemented using RSO
(Remote Spool Output) and is of course available to all Dprint clients. In addition, RSO can be used by openUTM users via a
special OLTP interface from UTM applications as a spool system for LAN printers.

Océ PRISMAproductionServer: Router
This Océ product includes all necessary functions for the operation of IPDS high-performance printing systems from different
manufacturers in miscellaneous mainframe environments. Its BS2000/OSD variant is a joint development by OPS (Océ Printing
Systems GmbH) and Fujitsu and is exclusively distributed by Océ.

New functions and further development of spool and print services
 Since October 2009 only the router version is offered for new shipments of high performance Océ printers, i.e. the modules
SPS and HP2AFP have been withdrawn from Sales. The print output (with compatible end-user interface) is standard via a
PRISMA production server connected to LAN. In particular, there is no direct BS2000 channel connection any more for high-
performance Océ printers. The conversion of existing SPS configurations to this model is supported by Océ via special
measures. As no printer driver in BS2000 is required any more, the Océ printer range is thus complete and supported without
any time delay by the previously required BS2000 porting. The printer administration is no longer in BS2000/OSD, but on the
PRISMAproductionServer, whose extended functions can also be fully used.
 The Remote SPOOL product RSO supports UTF8-capable network printers (specific example: Printronix P7000) as of
version 3.5 and as part of the overlapping Unicode project. The current version RSO V3.6 supports the protocol IPP V1.1 for
secure network printing whereby the data is optimally protected at transfer in LAN or Internet.
 A new user command CONV2PDF has been developed for the local SPOOL in order to convert one (or more) text files(s)
into one (or more) PDF files(s)



Other functions
The Spool Notification Service (SNS) is able to detect any event concerning any Spool & Print device or job, and launch
dedicated actions like e-mail, file logging or procedure execution. Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 6 / 14

The BS2000/OSD Print Services (Dprint in combination with local SPOOL and RSO as well as with Router) provide the
following customer benefits:
 optimum use of resources
 flexible cross-platform resource allocation
 high standard of security and availability
 increased productivity
 cost-saving post-processing

5.2.3 Data backup and restore
Exploding storage capacities in data centers pose challenges for the management of data and its resources. The trend toward
lifecycle-centric and system-controlled management of mission-critical data is the answer to these challenges. HSMS caters for
this trend on the BS2000/OSD platform.

Hierarchical Storage Management (HSMS) is the BS2000/OSD product for data backup and restore as well as migration on
the BS2000 platform. HSMS supports backup, archive, import/export and migration. As with all state-of-the-art backup products,
different backup methods (full backup, differential backup and incremental/partial backup) are available. Data can be saved to
different media, disk and tape. Data staging can be effected by copying save files from one backup medium (e.g. disk) to
another (e.g. tape).
HSMS offers excellent performance and is highly scalable especially for backups in configurations with ETERNUS CS. From a
user’s point of view a backup can be controlled by specifying service requirements through system-supported management
classes.

HSMS cooperates perfectly with BS2000/OSD-products like MAREN, ROBAR, AVAS and SHC-OSDand with Fujitsu ETERNUS
CS. Together with storage subsystems Fujitsu ETERNUS DX, EMC Symmetrix and CLARiiON and their software components
EquivalentCopy (EC), TimeFinder or SRDF and SnapView respectively, storage subsystems with high-end features can be
designed and run on the BS2000/OSD platform.

Following functions should also be mentioned:
 The availability of applications can be increased considerably if the DMS files used by the application are backed up using
the Concurrent Copy function. While Concurrent Copy is backing up a DMS file, the application still has read and write
access to the file.

 If the data that is to be backed up is located on ETERNUS DX, Symmetrix or CLARiiON CX systems, the data backup
process can be considerably improved by using the respective mirror function. The basic idea for the use of this function is
that the mirrored disks of the pubset are split at the required backup time and the backup data is then read from the mirror
disks.

 Online backup for databases (SESAM, UDS) with the HSMS/ARCHIVE/CCOPY products using the Symmetrix/TimeFinder or
CLARiiON/FibreCAT CX SnapView/Clone functionality. This enables the impact of the backup on the performance of the
database to be minimized. The split phase at the checkpoint is kept short, as with the existing CCOPY function. As with
SESAM, the Clone or BCV backup with HSMS is integrated in the DB backup functions.

 Shadow archives can be used as an easy means of automating the duplication of backup resources (e.g. for exporting to fire
archives).

 The capacity of magnetic tape cartridges has increased dramatically in recent years. HSMS can make optimum use of these
capacities, since the data from new backup runs can be appended after already completed backup runs on the same
magnetic tape cartridge.

 Storing backup data on disks is particularly suitable in situations where individual backup runs are able to produce only slow
data streams while the backup data is to be output on high-performance tape cartridges. The area on the disk is used as a
buffer in this case. Moreover, the backup data on the disks can be used as a fast cache for restore operations since there is
no need to mount a tape or wind to a specific position on the tape.

 The support of I/O parallelization via PAV or RSC results in considerable performance improvements. The parallel disk IOs
mean that the LTO cassette operation is not slowed down by possibly slower disk operations.

PAV (Parallel Access Volume) for disk I/Os on S servers
PAV allocates several device addresses (alias devices) to a logical volume. Seval I/O jobs can thus be set in parallel on a
volume.

RSC (Remote System Call) for disk I/Os on SQ servers
In conjunction with the RSC for bus disks a further tuning measure has been implemented: I/O parallelization: up to three
I/Os are now simultaneously possible on one volume/device; the serialization is then in the disk storage system.

The RSC funcion was transferred to the Linux-based carrier system X2000 V5.1 and is available with all SQ servers as
standard.
Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 7 / 14
HSMS and ARCHIVE parallelize disk accesses by using on S servers the feature PAV (Parallel Access Volume) and on SQ
servers the I/O parellelization in the feature RSC.
For PAV or RSC usage in ARCHIVE intervention is (only) centrally in the buffer administration and in reading/writing files
(with save/restore): Several asynchronous IOs are issued which then result with PAV/RSC in parallel execution and without
are executed sequentially (the I/O parallelization in RSC exists as standard on SQ servers). Furthermore, asynchroous
processing was activated when copying from disk-save file to volume-save file.


HSMS/ARCHIVE:
Supports file-multiplexing and PAV/RSC (up to 8 parallel data streams of
one file on several disks to one LTO-device; speed up by accessing disk-
devices in parallel)
Scalar 10k/i500/i6000

Figure 4: HSMS – Support of PAV/RSC functionality

New functions in HSMS V9.0B
 ARCHIVE/HSMS support of the OSD V9.0 NET storage support:
In the OSD V9.0 the data in NET storage also belongs to the scope of the backup of a pubset. Thus, HSMS/ARCHIVE are to
be extended to include functions for the backup/restore of files on NET storage.
This is why the data on the NET storage is in the case of backups always to be stored with the previous standard settings in
ARCHIVE and HSMS. However, the ARCHIVE user interface is not to be extended.
It is possible via extended HSMS interfaces to specifically select the backup volume with regard to files on local pubsets and
files on remote NET storage.
The selection according to NAS storage files is enabled for restore, and NAS storage is also offered as the target for the
restore.
 Extended search function for archiving:
After years of tape storage and the required reorganization and refresh cycles of tapes, the original assignment of archived
data to the save version is now obsolete. To support the organization and reorganization of long-term archiving a new search
function is offered in SHOW-ARCHIVE which performs the following:
- Evaluation (pattern search) of the comment entered in the user-information field
- Output of a list of all save-version ids and the user information

Outlook
Further development of HSMS will focus on the user requirements. Functional enhancements for the future version of HSMS
are planned in the following areas:
 Ease of use
 Performance improvements

5.2.4 Performance monitoring and trend analysis
Part of the design horizon of autonomic IT systems is the ability to adapt to different requirements and operational situations.
Continuous monitoring of resource utilization and performance behavior is necessary in order to assure this adaptability. The
most important sources of information for this are the objective data provided by system performance measurements. While
round-the-clock logging of a few central metrics is sufficient for trend analysis, very detailed performance data must usually be
provided in order to perform a complex bottleneck analysis. In clustered or distributed systems, it must be possible to monitor a
number of BS2000/OSD systems simultaneously from a single central point so that any degradation of performance can be
instantly detected during online operation and the causes analyzed. Deviations from normal operational behavior can be
recognized immediately by threshold value checks on relevant metrics. For reporting purposes, meaningful graphics are
essential for visualizing the statistical performance data.
PAV/RSCSummary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 8 / 14

openSM2 (BS2000/OSD) records all the relevant system performance data. The scope of performance data logging can be
adjusted to suit current requirements. The performance data is made available for online monitoring and can be saved to
monitoring files for later analysis. The ANALYZER tool enables monitoring files to be analyzed and the measured data
presented graphically on a PC. The INSPECTOR tool provides simultaneous online monitoring of multiple BS2000/OSD
servers, delivers a graphical presentation of the measured data and issues alerts if there is a threshold violation. When
combined with the other members of the openSM2 product line, openSM2 can be used to implement enterprise-wide
performance monitoring of server and storage systems by enabling Solaris, Linux, Windows, Xen and VMware ESX Server
systems and ETERNUS DX, Symmetrix, CLARiiON and FibreCAT storage systems as well as BS2000/OSD systems to be
monitored online using threshold values and alerts. Additionally SNMP-enabled systems can be monitored.

openSM2 (SNMP) collects selected data of the following groups of the SNMP MIB-2:
 IP
 TCP
 UDP

These tools offer two key benefits. Firstly, the budget owner has an overview of medium- and long-term trends at any point in
time and so can control the allocation of resources in the most effective way. Secondly, system administrators have detailed
information at their fingertips, enabling them to alter system settings to assure optimum use of resources and compliance with
service level agreements.


Figure 5: openSM2


The current version openSM2 V9.0 (BS2000/OSD) offers the following enhancements:
 Monitoring of openFT applications
 Improvement of analysis functions
 Enhanced support of RSC-IOs
 Support of SQ Business Servers (VM2000 measurement data)

The current version openSM2 V9.0 for open systems offers the following enhancements:
 Agent-less monitoring of Linux and Xen systems
 Monitoring of openUTM applications on Solaris and Linux systems
 Monitoring of devices supporting SNMP
 Simplification of the installation and configuration
 Improvement of analysis functions
 openSM2 (Storage) V9.1 monitors now additionally ETERNUS DX Storage Systems

For the next version openSM2 V10.0 (BS2000/OSD) the following enhancements are planned:
 Improvements of montoring and analysis

Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 9 / 14
5.2.5 Management of intelligent disk subsystems
SHC-OSD (Storage Host Component for BS2000/OSD)
Maximum data availability is one of the indispensable preconditions for modern business processes. Depending on the sector,
application structure and duration of the downtime, a temporary interruption to access operations (scheduled or unscheduled)
can very quickly lead to serious image damage and loss of sales revenue. Data mirroring, either on the local system or remotely
over quite large distances, can provide the remedy.

The product SHC-OSD is the BS2000 host component for disk subsystems. SHC-OSD provides information services and
commands for controlling data mirroring in BS2000/OSD.
The disk storage subsystems supported by SHC-OSD provide the basic functionality for data mirroring:
• Fujitsu ETERNUS DX with the replication functions EC (Equivalent Copy) and REC (Remote Equivalent Copy)
• Symmetrix with SRDF (Symmetrix Remote Data Facility) and TimeFinder
• CLARiiON CX with SnapView (an appropriate local data mirroring functionality for TimeFinder)
The storage systems Fujitsu ETERNUS DX400, DX400 S2 and DX8000 are released for operation with Fujitsu BS2000/OSD
business servers of the S and SQ series. The release of the models ETERNUS DX 8000 S2, also for the servers of the S and
SQ series, is planned for the 1st quarter of 2012. Symmetrix disk subsystems can be connected to all BS2000/OSD Business
Servers. CLARiiON CX disk subsystems are supported on BS2000/OSD Business Servers of the SQ series and SX series.

The following customer benefits can be achieved by using SHC-OSD:
 Failover/failback operations can be coordinated directly from the BS2000/OSD system. Intervention by an engineer is not
necessary.
 Application availability can be improved substantially, since creating copies and synchronization are performed in parallel
with production operation.
 Customers are provided with great automation potential, since this command set offers optimal integration into procedures.
For easier process control, commands are also provided in a synchronous form. This means that all related actions, e.g. copy
processes, in the storage system must be completed before the next command is started.
 System integration ensures that no inconsistencies can arise at I/O level when mirrored volumes are split off. Parallel data
resources can therefore be created “on the fly” for recovery-capable applications.
 The BS2000/OSD can be used as the administration center for all disks, because even non-generated or deactivated units
can be displayed and administered.
 The abstraction of the hardware functions by SHC-OSD results in the simplification of the introduction of new firmware
versions of the supported storage systems.



Figure 6: SHC-OSD

In the Symmetrix case SHC-OSD uses EMC’s strategic open interface SYMAPI (Symmetrix Application Interface) for EMC-
developed software and partner products. SYMAPI is a strategic EMC interface which is also used by many other systems. So
maximum investment protection is guaranteed. EMC is also treating the compatibility issue as a top priority for the further
development of SYMAPI.
Administration with SHC-OSD is in the case of ETERNUS DX implemented fully via StorMan. The StorMan component runs on
the external Windows Management Server of the ETERNUS DX systems. StorMan uses the embedded ACCC-API, which also
runs on the Windows Management Server of the ETERNUS DX systems. StorMan also needs the Fujitsu SMI-S provider. In a
configuration with ETERNUS DX400 or ETERNUS DX8000 the Fujitsu SMI-S provider runs on the Windows Management
Server. In a configuration with ETERNUS DX400 S2 the Fujitsu SMI-S provider runs on the ETERNUS DX400 S2. Summary  Issue: November 2011  BS2000/OSD Scope – Chapter 5 System management  Use: External Page 10 / 14
In the case of CLARiiON the SHC-OSD uses the SMI-S functionality in the CLARiiON CX, which is provided via the SMI-S
provider from EMC. The SMI-S provider from EMC runs on SQ servers in the X2000 part, in the case of SX servers it runs on an
external Linux or Windows server. A StorMan component has also been implemented on BS2000/Posix. StorMan implements
storage functionality on the basis of SMI-S functionality and thus encapsulates the specific properties of CLARiiON CX.

New functions in SHC-OSD V9.0 (current version)
Support of the ETERNUS DX
The replication functions for storage systems from Fujitsu (ETERNUS DX410/DX440, DX410 S2/DX440 S2 and
DX8400/DX8700) can now for the first time also be controlled from BS2000 with the current version of SHC-OSD. Furthermore,
information and monitoring functions are also made available for the ETERNUS DX systems as of this version. Simultaneous to
their release for BS2000/OSD, the support of ETERNUS DX DX8000 S2 Disk Storage System is planned for the 1st quarter of
2012.

Information functions for ETERNUS DX
The offered information functions for ETERNUS DX cover general information about the storage system and storage volume
configuration (BS2000-specific details about MN, VSN, device type, as well as the appropriate storage-system-specific
information, such as serial number, logical volume number, RAID level, status, configured capacity, etc.), and information about
local mirroring with EC and remote mirroring with REC (details about MN, VSN, original and mirror disks, the logical volume
numbers, type and serial number of the controllers, mirroring status, etc.). The information functions for ETERNUS DX storage
arrays are mapped on existing interfaces.

Event Monitoring
SHC-OSD V9.0 also offers for ETERNUS DX systems the event monitoring that is offered for Symmetrix systems and
CLARiiON CX in the form of BS2000 console output in the event of status changes for the storage subsystem and for individual
volumes, also with regard to active data mirroring.

Local mirroring (clones) with Equivalent Copy (EC)
EquivalentCopy (EC) provides local, continuous mirroring on a volume basis with separable mirrors. A local mirror volume of
identical size is allocated to an original volume and after initial synchronization is also carried as a mirror. A maximum of 16 EC
mirrors are supported for one original by SHC-OSD. Multiple mirroring is supported for a maximum total of 32 mirror volumes.

Synchronous remote mirroring with Remote EquivalentCopy (REC)
It supports mirroring on a volume basis with separable mirrors between 2 or more ETERNUS DX systems. A mirror volume of
identical size is allocated to a local original volume in the remote ETERNUS DX and after initial synchronization is also carried
as a mirror. At present a maximum of 16 remote ETERNUS DX systems can be connected. Multiple mirroring is supported for a
maximum total of 32 mirror volumes. A maximum of 4 simultaneous REC mirrors are supported by SHC-OSD for one original
(concurrent remote copy).

New Symmetrix functions
Virtual Provisioning: Selection of logical volumes via pool names
To enable simple administration of virtual provisioning the selection of the logical volume numbers by (thin) pool names is
provided for the commands /SHOW-SYMM-DEV-CONF and /SHOW-STOR-DEV-CONF .

Functions for the administration of the RA groups
The support introduced with SHC-OSD V8.0 of Dynamic SRDF (START-REMOTE-COPY and STOP-REMOTE-COPY) also
requires for the option of dynamic SRDF functions for the administration of so-called RA groups, which define an SRDF
connection between two volumes of two Symmetrix systems. For this purpose commands (ADD/MODIFY/REMOVE-
SYMMETRIX-RA-GROUP) are provided for the administration of the RA groups and their allocation to local and remote
directories.
The functionality is particularly important for the use of SRDF/A, because administration here does not take place at volume
level, but via the RA groups.

DataMobility configurations on the basis of SRDF/AR
In addition to the already supported synchronous SRDF mode (SRDF/S), DataMobility configurations on the basis of SRDF
mirroring in asynchronous "adaptive copy" mode are supported with the correction level for SHC-OSD V9.0, which is planned for
Q1/2012.

Outlook
New functions in SHC-OSD V10.0 (planned release: 2nd half-year 2012)
The support of local mirroring with snaps and virtual provisioning for ETERNUS DX systems is planned for the follow-up version
of SHC-OSD (V10.0), to be released probably in the 2nd half-year of 2012. The support of asynchronous remote mirroring is
planned for a further follow-up version.

SCCA-BS2 (Storage Control Center Agent for BS2000/OSD)
SCCA-BS2 implements the integration of BS2000/OSD and the connected Symmetrix subsystems into the EMC ControlCenter
product family, which provides general functions for visualizing, monitoring, automating and providing storage resources in a
heterogeneous storage and server environment. SCCA-BS2 is based on the strategic Symmetrix interface SYMAPI and on the
SYMAPI server, both of which are provided by EMC Corporation.

Be the first to leave a comment!!

12/1000 maximum characters.