input structure for executive summary by sanmelody


									Ohm                                                   Kurzinfo                                      20.03.2011

                                                    input + structure for executive summary
topic: registered status, recommendation                                                effort/ cost, time                                benefit
Upfront Effort
Prior to immediate implemention of any recommendation, the following details should be investigated and specified as a follow-up to the study according to the
suggested roadmap (WBS).
Besides, most of the recommendations require some detailed design work before actual implementation.
This is, why there will be a higher initial effort when trying to apply the recommendations for the first time, while the longterm overhead will decreas.
Identify layers of applications on infrastructure                                          contribution from all ESO-sites         avoid timeconsuming adhoc
As a baseline for any further activities, the study recommends to first agree on a             and IT-units needed, i.e. about         activities to cope with urgent needs
common view of the applications within ESO and the underlying infrastructure. This             5-8 MD´s per person for about           for changing or enhancing IT-
should be a layered overview including actual status and desired target.                       10 persons                              infrastructure, which the IT-units
This overview should then be turned into a commonly agreed description of "the                                                         could not foresee, when operating
ESO-standard" for IS-components, i.e. regarding                                                                                        with little integration to the
 equipment, tools                                                                                                                     applications
 commercial SW                                                                                                                     essential for a realistic sizing and
 freeware                                                                                                                             scoping of infrastructure
 suppliers.                                                                                                                        essential for understanding of
The standard-documentation should describe                                                                                             purchase-needs, i.e. the basis for
 the frame within which something is perceived as standard-compliant                                                                  approvals
 the range of allowed deviation and exceptions
 the process of how to change the standard-frame

This acitivity is best setup as a real project with about 2 representatives from each
site/ IT-unit. It should follow a stepwise approach for the most important
applications and underlying infrastructure at all sites first (VLT-control, DFS+
reduction, ERP). The findings of the study can be used as a basis. Best would be to
document the results within the inventory.
Once, the common view and a first standards version are established, measurable                 contribution from those ESO-         help aligning the infrastructure with
targets for infrastructure and maintenance-services can be defined.                              sites and organizational-units        the actual application-needs
The applications´ workload should be documented, both, as needed in short- and
                                                                                                 needed, which provide/ run           facilitate communication to users
                                                                                                 applications; this is not             and setting user-expectations
longterm, in terms of transaction-volume, computing-times, response-times,
network-througput. Similarly, the applications´reliability should be documented, in
                                                                                                 necessarily the IT-departments,      demonstrate the success of
                                                                                                 but rather those representing         improvements when comparing to
terms of MTBF/ MTbSI and MTTR, i.e. the applications should require SLA´s to be
                                                                                                 the user-expectations                 measurable results
kept by the infrastructure and the respective services.
Possible disasters (concerning workload, reliability) and their impact (also in terms           some contribution in                 evaluate the success of contracting
of SLA´s and with respect to research) should be identified in advance, so that a                discussions also from IT-

Sibylle Ohm               D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc          Seite 1/13
Ohm                                                  Kurzinfo                                20.03.2011

topic: registered status, recommendation                                            effort/ cost, time                             benefit
further specification of how ESO wants to cope with these is feasible. There should   about 2-3 MD´s for 15 persons
be an ESO-wide understanding of prioritization in case of disasters.                  guessed as initial effort
Statistics to measure the actual performance of infrastructure and services should
be defined and implemented. Such SLA-statistics should cover ticket-handling,
support-delivery, networking (SNMP), applications and management-reporting, and
run during well-defined time-periods. They help in monitoring for approaching to
capacity-thresholds so that bottlenecks can proactively be avoided.

The following suggestions of improving the ESO-wide infrastructure and consolidation of the huge computer-park at ESO to the actual needs, are based on what
the study found out when doing the first steps in the "upfront effort" mentionned above. The recommendations are on HW, SW and networking.
Actually, there it is common practice at ESO to equip each instrument with            This redesign is a major effort, i.e. it  The same infrastructural concept
dedicated workstations, ending up in a large lot of computers at the observatories, should be carefully specified in             applies to all instrument-
all of which must be maintained separately. The workstation HW is very old            advance. Any change to the control-        workstations. There is much less
(average 8 years or more for both observatories).                                     area requires longterm and thorough        dedicated maintenance-attention to
ESO should verify whether this HW-situation for telescope-/instrument-                test-activities, before it can become      pay than for single workstations.
control-workstations, which gets even more complex with each new instrument, operational.                                       Instrument-dedicated HW,
should not be generally redesigned. There are several options to consider:            This may well take a period of 1 or 3      operating-system and connectivity
     Rack-mounted blade-technology, best with central disk-, RAID-, backup-          years.                                     is saved (though it must be
      facilities, i.e. the approach of dedicated "workstations per instrument"                                                   assured, that central components
      eventually is kept, using small boxes, independent from each other, with high                                              are powerful enough). Under-
      scalability. Blade-technology can easily be equipped with load-balancing                                                   utilization can be more easily
      (depending on the applications they run), so that this could be a step into a                                              avoided.
      more integrated direction as well.                                                                                        Overall planning and overview of
     A parallel system (MPP) set-up in a way, that problems with one instrument                                                 an observatory´s control-area is
      do not affect the others.                                                                                                  faciliated; this especially affects
The HW-redesign may go along with the recommended stepwise and site-                                                             The scoping- and integration-work
comprising migration from HP to Linux.                                                                                           moves to SW-side from the HW-
This should investigate pro´s and con´s as well as necessary planning of                                                         related issues.
migration-tasks with all parties (the study again can serve as a baseline for
initiating this migration): PAO-mgmt and PAO-IS-groups (SW-group, PAO-IT,
data-handling), LSO-mgmt and LSO-IS, GAR-TEC incl. TWS and OSC-IT, GAR-
DMD, GAR-INS, GAR-VLT/TSD, SGO-IT and SGO-mgmt. The following
migration-steps are suggested:
 first replace HP by Linux within the DFS (to test in office-lab-environment)
 replace within telescope-control-area but first the servers running Maximo,
    Autorep, Remedy-apps, etc.

Sibylle Ohm              D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc    Seite 2/13
Ohm                                                   Kurzinfo                                        20.03.2011

topic: registered status, recommendation                                                     effort/ cost, time                               benefit
 replace within LSO for the telescope-control-area workstations, first 1 selected
    telescope, then the other telescopes
 replace within PAO for the telescope-control-area workstations, first 1 selected
    telescope, then all telescops
At PAO there is not enough direct-access-storage for data-reduction available.                  cost for extension of storage and         Observations can become more
At all Chile-sites, the data-reduction computers have become outdated and do not                 computing power                            "interactive", when astronomers
fulfill the user expectations.                                                                  time needed to commonly agree              can analyse observations of the
The workarounds, at least those at PAO to cope with the complex workaround of                    on the ESO-wide "standard-                 previous night in order to be
accessing archived data, should not be necessary:                                                reduction-infrastructure" and the          advised for the next observations;
Storage at PAO for scientific analysis should be enhanced, LSO would need more                   "standard-proceeding in further         Good data-reduction capacity
computing power for reduction machines.                                                          extending it".                             allows for direct checking of data-
                                                                                                                                            quality close to the telescopes.
                                                                                                                                         ESO´s "customers", expect better
                                                                                                                                            reduction functionality also on the
                                                                                                                                            observatory-site, (and when
                                                                                                                                            comparing to other observatories)
                                                                                                                                         Scientists (external, ESO-internal
                                                                                                                                            operators) and IT-SysAdmins
                                                                                                                                            spend less time with workarounds
There are a couple of "smaller" HW-related issues, which may be of less                         a few MD´s (about 10-20) for the       Users (scientists, engineers,
importance, but for which common solutions can be more quickly be found and                      commonly agreed solution               SysAdmins) spend less time in copying
implemented, such as the a useable replacement for X-terminals or the issues                     design per issue,                      with unsatisfactory workarounds or
with running Linux on Dell-HW. For each of such issues, there should be a                       some more MD´s plus evtly.             even error-situations.
responsible person nominated to lead the topic to an agreed solution and to                      small HW-expenses for
suggest a schedule for the respective implementation.                                            implementing the solution (either
                                                                                                 ongoing or at once)
At ESO Macintosh´s are prefered by some users for different reasons.                            Macintosh-purchasing is about             facilitate publishing;
 A user just has a personal preference.                                                         twice as high than for another            avoid dual-boot in exceptional
 For publishing reasons, Mac-applications would be more appropriate.                            desktop-type (MS or Linux). TCO            cases
 With Macintosh dual-boot-configurations are unnecessary.                                       statistically looks better for Mac´s      users prefering Mac´s are less
 Within the astronomer community, Macintosh is common with increasing                           than for MS-PC.                            efficient on other environments
    popularity, though Linux-usage is also increasing.                                          Additional know-how about Mac-
As Apple is a niche-player, the future of which is more uncertain than that of Linux             SysAdmin must be available at
and MS, it should not become ESO´s strategic platform. ESO should document                       ESO.
the allowed reasonable exceptions, the process for providing a user with                        ESO-standardization becomes
Macintosh-environment and the related maintenance-service, the user can                          more complex

Sibylle Ohm               D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc             Seite 3/13
Ohm                                                              Kurzinfo                                           20.03.2011

topic: registered status, recommendation                                                                   effort/ cost, time           benefit
file-sharing + account-handling (context with backup+archiving)
(2 types of store-spaces for astronomers: a working-space of 100 GB and a
result-space of 20 GB which is daily backuped (office-for-science BLeibundgut))

     consolidate HW for this as well
      file-server at SGO
 introduce ADS, CCM at LSO
 media-concept: Setup and implement a concept for handling of different
      storage-media (tapes, CD´s, DVD´s, spinning-disks, others) and their
      read/write-devices, which should not be scattered around at any workstation
      (as Linux-workstations usually are lacking all needed slots) but more centrally
      organized at known major servers. (e.g. USG, OSC)
This goes along with archiving and backup-concept.
 groupware (calendars, conf-rooms-booking, secretary-communciation, PDA-
      connectivity), share observations,
 which files and file-types
 by which subdirectory-structure
 through which technological means (Unix- or MS-oriented, simple file-system,
      intra-/ extranet, document-archive for ERP/ other, other workflow-application
 policy for user-types+ access-rights, tear down of left users

LDAP (ADS) ist prima, Linux/ unix fehlt noch; ESO-IS should establish one unique
source, am besten ein LDAP (nicht eMail oder ERP, weil da nicht unbedingt die
visitors im ERP stehen, und denn braucht man wieder eine Ausnahmeregel)

backup                                                                                                                               avoid loss of data (auch wenn das
backup-means + services at SGO, LSO, PAO                                                                                              soo häufig noch nicht vorkam, weil
document what contents is how to be backuped at what sites for what reasons                                                           ja nu das controls-zeugs über
Specify a plan and documented concept of growing storage-capacities at all sites in order to cope with
the ever-growing data-volume. (based on "upgron effort" finding)
                                                                                                                                      klaren DFS-Mechanismus
 That plan has to align with projected usage of instruments.                                                                         weggeschrieben wird.
 Unify the high diversity of storage-means (single disks, all RAID-levels) at all ESO-sites, eventually                             reduce time-consuming, manual
    foresee centralized storage-means (SAN, NAS) for file-servers at each Chile site respectively,                                    error-prone, demotivating restore-
    instead of having local-computers-attached-storage only.                                                                          operations
                                                                                                                                     Ad-hoc buying of additional disks is avoided.
und hierhin, was wir zu infrastructure+ service tatsächlich vorgeschlagen haben zusammengefaßt                                       approvals of purchase-requests

Sibylle Ohm                     D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc                     Seite 4/13
Ohm                                                              Kurzinfo                                         20.03.2011

topic: registered status, recommendation                                                                 effort/ cost, time                            benefit
                                                                                                                                                    can become nachvollziebar
                                                                                                                                                   less person-dependent, better
 LAN: HW OSI 1,2: GAR ok, LSO a mess, PAO ok (apart ATM), SGO mittel?
 VPN at SGO-GH is not (yet) available;

Specify LAN/WAN-load generated by app´s+users
 actual
 and regularly (once a year) upcoming expected traffic in future
focus not only on WAN, but also on a common LAN-concept over all sites, i.e. 1 approach for office-
tasks, 1 approach for control-area-tasks

WAN redesign-> impact on LAN according to applications;
subnetting and eventually even different internal VPN-zones,
zoning: secret, internal, external, internet zone
zones on top of one common underlying VPN
while replacement for satellite and leased-lines (incl. clarify actual work on
   redundant micro-wave equipment does not seem to be needed
   satellite devices should be kept, provider contract should foresee means to get back to satellite-
    operations within tbd SLA-time
   micro-wave should link to the two Chile-rings (one main, the other standby)
   requirements for traffic-volume over micro-wave should not only consider the theoretical maxima
    but also the actual average usage

Specify SNMP-usage (Openview)

Introduce/ Implement these spec´s, stepwise


     stepwise implementation
    the general good idea of involving high-skilled specialists for a limited time to
     solve bottlenecks, problems and suggest future concepts, specifically applies
     to networking
eMail                                                                                                    users do their own sysadmin on their
clients, backup, server-consolidation ESO-wide                                                           client-varyity

Sibylle Ohm                      D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc                  Seite 5/13
Ohm                                                               Kurzinfo                                            20.03.2011

topic: registered status, recommendation                                                                     effort/ cost, time     benefit

   The document-archive coming along with the ERP-system, has specific requirements to eMail-
   Regarding eMail-archiving, there are little legal requirements to ESO re. interfacing outside parties,
    but eMail-archiving should be taken as an ESO requirement for its internal backup.

shortterm improvements
    prescribe 2 clients (1 Unix, 1 Windows+ Mac), and define and publish allowed interfaces
    (SMTP und IMAP) for any other clients
     pull Spam-functionality out of eMail-server-HW onto separate HW
     improve Spam-filter-mechanism (new SW)
     introduce eMail backup
renew eMail HW-configuration
        eMail-servers at SGO and at GAR only

define handling of accounts, mailing-lists and routing
    responsibilities for managing accounts and mailing-lists
    document policies (incl. parametrization+ configurations) for establishing and changing
       accounts+ mailing-lists
    implement stable automated synchronizations (from account-definition master, e.g. the ERP,
       and between the eMail-servers)
    establish and publish backup-rules
   LDAP yes/no

review eMail triggering automations
cronjobs et al should not all start at the same time

define eMail-similar ESO-standards
e.g. on chat (urgently needed by virtual SW development teams) and on diaries/ calendars (urgently
needed by more administrational personnel)!!!
on discussion-forums

specify application-solutions
    for Navision/ ERP
    for several Remedy-applications
    BCSW
    P2PP
    specific eMail-handling for "external" telescope-institutes on ESO-sites, e.g. former SEST at

Remedy (and other) applications

Sibylle Ohm                      D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc                      Seite 6/13
Ohm                                                               Kurzinfo                                        20.03.2011

topic: registered status, recommendation                                                                 effort/ cost, time     benefit

list existing application+ model what these cover
            also consider applications providing similar functionality but not implemented through
            Remedy, e.g.:
            - Maximo at PAO,
            - environment-monitoring-SW at PAO,
            - keycard-application at PAO,
            - Swrty-tracking-tool of PAO-IT should specifically be considered (looks very professional
            and new, none-Remedy-development for tracking-tools could take this as a basis)
            - the whole bunch of SGO-specific applications, such as ActualTrip, reporting, vacation,
            guest-tracking, taxi- and meal-control, AQCT for PABX, etc.

decide future applications coherence (data-model)
decide on build or buy ERP, selbst entwickeln,        code-generator (Cebit)
transfer to operation
agree on process for contin.improvment

web-applications+ internet-Auftritt
impact of intersite-communication-plan
ESO std-web-technology
rules on web- and intranet-contents
consolidate web-servers (HW, operating-systems, Web-Server-SW, Servlet-containers, ASP-
interpreters, scripting-interpreters, application-servers, etc.)
decide on redesign look&feel (what would be the effect to ESO if the look& feel becomes more
Services, Staff/ Organization
With the current organizational and process-related conditions at ESO,
improvements in service-delivery can be only marginally (so there is hardly any
big cost-blocks to identify, which could easily be reduced) and current
shortcomings can hardly be avoided by the respective team.
Precondition to optimally deliver IS-services is a completely different IS-
organization as it is currently.
hotline sollte extern sein; besser als in OSC-IT mit interner geht nicht;
es reicht gerade, wenn support-Aufgaben ständig die "Projektaktivitäten" stören
service-delivery context                                                                                 -                      -
activity-blocks are specific to the respective supported or maintained IS-

Sibylle Ohm                      D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc                  Seite 7/13
Ohm                                                             Kurzinfo                                      20.03.2011

topic: registered status, recommendation                                                             effort/ cost, time                         benefit
environment; and specific to whether office-environment or telescope-

         user-facing                                 HW/ Oper.System
         hotline, support
                                                      rel. database
                                                     standard MS SW
         install, upgrade, monitor,                  ERP/ Navision
         coninuously improve                        intersite networking
                                           Office                        Telescope

hotline+support                                                                                      Vorbedingungen für externe hotline:        SLA´s in peaks, tolles-knowhow,
define incident-categories + dispatching-process                                                    - services müssen definiert sein,          in Chile über Nacht auch, user
define all IS-environments which should be covered by support and which level                        - Gruppen und dispatching muß klar sein,   müssen nicht ständig wissen,
define prio´s, SLA´s, escalation (incl. disaster)
define solving-groups (according to the ticket-categories and dispatching, and define the solving-   - remote-Zugriff muß eingerichtet sein;    wen sie dieses/ jenes fragen,
group´s service)                                                                                     -> erst dann kann die Ausschreibung dazu   sondern melden sich an 1 SPoC
define complaint-handling                                                                           passend gemacht werden                     wenn sie ja uch nicht unbedingt
define interface incident - problem mgmt                                                                                                       wissen, woher ihr Problem
define requirements to ticket-application                                                                                                      eigentlich stammt
define requirements to knowledge-DB

information policy; common rules for information to users; clear communication;
rules for feedback from users;
rules for keeping suitable with ESO-general-strategy s.below (diese 3 gehören ja
knowledge/ confidence / Erwartungshaltung des users verbessern, andere user-
dept´s treten sie in den Hintern, weil sie priorisieren müssen, wie sie ihre wenigen

Sibylle Ohm                     D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc               Seite 8/13
Ohm                                                            Kurzinfo                                     20.03.2011

topic: registered status, recommendation                                                           effort/ cost, time     benefit
Leute einsetzen, z.B. 20 ECF-Leute kann er nicht vernünftig supporten, weil er
sich eher auf scientists konzentrieren
SysAdmin-related tasks (infrastr.availability, std-SW, ESO-specific SW)
evtl für die Liste der recommended services noch jeden einzeln auf wichtigkeit hin

specify standards for delivery of maintenance-services; zumindest hier listen als
Einstieg zumindest in eine solche Spez der standard-services, wenn sie sie schon
so nicht übernehmen wollen

gerade dringende:
ERP/ Citrix maintenance concept (und Citrix _Notwendigkeit hätten sie schon
vermeiden sollen; möglichst nicht ausdehnen)
urgently dedicate 2 ESO-persons to Navision, even if these have to be new hires
1. process
2. means and tools
3. suppliers
 Cisco (Checkpoint-FW?)
 Dell
 EMC (Storagetek, Seagate oder andere?)
 TPP statt Sun, HP
 TPP für Linux-Bastelei statt Dell

Last but not least projects differ from the organizational structure in that they are
only temporary. Role, task and organizational integration of project managers
must be clearly defined. A project’s location (whether a SW-development- or a
SysAdmin-project) can be at any site, (so that e.g. people rotating between the
Chile observatories would be assigned to one dedicate site for the duration of the

am Beispile der list of OSC-IT projects
Die Projektliste 1 Seite Papier von GFilippi, z.T. ganz gute Themen

Sibylle Ohm                     D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc             Seite 9/13
Ohm                                                           Kurzinfo                                     20.03.2011

topic: registered status, recommendation                                                          effort/ cost, time                            benefit

incorportate IS into ESO-strategy; (regularly get updates of ESO´s overall
derive IS-forecast accordingly;
regarding HW, SW (incl. applications), networking (all 7 layers incl. security),
support + maintenance-services, knowhow+ staffing+ organization
e.g. growing percentage for service-mode-observing requires increase of supporting team-size

Standardize, Change-mgmt,                                                                         to avoid that users wonder about wasted
                                                                                                  money (memory-sticks, HW sized twice as
Of course, the initial standardization effort will require a higher level of personnel-           needed, licences of Skisoft-components        service-delivery-time, -cost and
utilization at first in order to reduce superfluous overhead in ESO-wide service-                 automatically extended without                staff-motivation
delivery in the long term.                                                                        investigating future necessity, wireless,
                                                                                                  WAN, prioritization of HD-tickets, when a
standardization of services across sites, across divisions                                        ticket is worked on/ expected to by solved,
 using the same technology, default equipment config                                             backup-proceeding, ...);
    (incl. printers, file-sharing+ group-work; RDBMS, etc.)
 for all service-delivery
 a method for recording and working on requirements to ESO-IS from
    research/ engineering
 establishing a regular checking and continuous improvement process of IS-
                                                                                                                                                increased know-how-level; less
document                                                                                                                                        person-dependent
 the standard itself
 frame for deviations
 process for continuous improvement
                                                                                                                                                  similar processes to be
benefit of ISO-9000-cert.?                                                                                                                         integrated into one process-
idea MCullum: regular lunch-time presentations                                                                                                   finding out synergies of
understand and actively advise users (understand the user-request and suggest                                                                      repeated activities
appropriate choice, instead of asking the user to choose out of 3 standards);                                                                    reuse or even use in parallel
 SW-dev-tools sysadmin should not be left up to the div. SW-dev-groups, but                                                                       of IT-equipment
    agree on ESO´s future std-tools and have IT-Sysadmin´s installing/                                                                          benefit from IS-persons’

Sibylle Ohm                    D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc            Seite 10/13
Ohm                                                     Kurzinfo                                        20.03.2011

topic: registered status, recommendation                                                       effort/ cost, time                               benefit
    maintaining the tool-basics (e.g. UML-tools, CASE-tools, test-tools,                                                                        knowledge for several purposes
 practicable de facto hardware and software standards defined by OSC-IT,
    replicated to other sites (e.g. PC-HW-config, ADS, router-config, CCM,
    Jumpstart, Kickstart, inventory)
 procurement: standards/ suppliers/ licencing, process
    Dell, Cisco, emc2 ok;
 standardize on printers todo

Then ESO should
 document the ESO-wide (sites, departments) standardized services as
 operate according to that documentation,
 measure the success of the service-delivery (with the help of SLAs)
 and continuously improve the services.
The study suggest the following structure for service-description:
- What (is done)
- Why (is it done)
- Skill (needed to do it)
- Technology (needed to do it, and environmental technology)
- How (is it done: steps, means, etc.)
- Success Criteria (SLAs)

Only when services become measurable, users have a basis for rating the
service-level of the overall organization

Initiate internal charging
for users´ backup
for calls during standby

The current IT-organization at ESO results in having major IT-related tasks                    Current organizational and process-related       more stable and predictable
scattered across several teams with little (or difficult to extend) cross-site and             conditions require big efforts for only slight   utilization of people
cross-division cooperation. Even worse, users in vain expect the same,                         improvements.

Sibylle Ohm                 D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc            Seite 11/13
Ohm                                                    Kurzinfo                                        20.03.2011

topic: registered status, recommendation                                                      effort/ cost, time     benefit
standardised IT-services, irrespective of delivering site or department.

s. manual drawing about summary-concept staff-orga
In addition, for each site, there should be one dedicated person from one of the
IS-units, working at that site, who is the site’s central coordinating contact for any
Besides, several non-IS divisions shall install IS-key-representatives.

major organizational restructuring of all of these local (with respect to space                                      ansonsten fehlt die
and department) activities so that better integrated as the one unique ESO-IS                                        Entscheidungsfähigkeit, welche
The study describes, how it developed its recommendation for the structure of the                                    user-requests "I hätte gerne"
new IS-organization, so that ESO might discuss and end up with a different result-                                   überhaupt sinnvoll sind, es wird
organization when applying the study’s approach.                                                                     dann halt gekauft
reorg based on recommended services

Approach: these major task-blocks are then first mapped to the ESO-site and
secondly mapped to the non-IS-, i.e. the user-site-organization, in order to
analyse the impact for the organisational structure
assess Merge of VLT-SW for Alma and PAO/SG
constraints are contradictory:
 take as few internal ESO employees as possible, i.e. no ESO-specific
    knowledge, high cost, tends to end up in body-leasing.
 reduce cost, i.e. increase the number of Chile-employees instead of
    European contractors
get most qualified persons onboard, i.e. increase the number of ESO-employees,
high cost

und wie bei services auch hier die orga-units einzeln auf wichtigkeit hin
utnersuchen, zumindest auflisten
jedenfalls die Trennung in MS- HW+ Anwendungen und Unix-/Netzwerk-
HW+Anwendungen und die Trennung in wenige interne leads/CR-/Projekt-Leute
und mehrere externe low-level-sysadmins ist schon ganz sinnvoll; - Mischung:
operational und exteranl zur unterstützung; die lernen dann auch+ identification
und in den Telescope-areas nur die operations abdecken;
-> also ein anderes team, das nicht site-dependent ist;

Sibylle Ohm                D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc            Seite 12/13
Ohm                                                   Kurzinfo                                        20.03.2011

topic: registered status, recommendation                                                     effort/ cost, time     benefit
zusätzlich nehmen sie zeitweise andere externe Spezialists für ein solches area
wie CMM, ADS, networking, und incl. know-how-Transfer; das ok!

aspects of a matrix-approach in its reporting

current synergies of having IS-services very closely related to business, such as
the SW-development for engineering or science respectively, are not lost

If ESO has to extend the outsourcing contracts, because it needs more time to
decide on its organizational restructuring, it should consider:
 for Terma: have the right to view the CV’s and select the people, as with
    common body-leasing
 for Serco: have activities been done by less expensive outsourced Chile-
    people instead of a high-price-company closely related to European
 for ADP: get MS- and network-skilled people on board, otherwise the ADP-
    approach can be helpful in doing so while covering the pure body-leasing.

in recommendation für hotline-service und HW-Tel (hatte ich zumindest bei housekeeping)
 telefony
     regardless of location, helpdesk should always be accessible via the same phone-nr
     users also want to keep their phone-number within all-ESO, irrespective of their actual location

Sibylle Ohm               D:\Docstoc\Working\pdf\18042fbb-6895-422b-83d6-281eafee2d81.doc            Seite 13/13

To top