Simultaneous access to a project in the Enterprise (SQL) version

[Patrick]

I'm an IT manager at a large chemical company. I'm evaluating the Enterprise
(SQL) version of MS Project 2003 as a possible alternative solution to
Primavera for our maintenance, turnarounds, and capital projects
organizations. I have a few high level, functional questions:

1. Can multiple projects be set up within a SQL DB?
2. How many simultaneous users are allowed to work within a project? within
a DB?
3. Can resources within a given project be assigned as project specific, or
are resources global within a DB?
4. How good is the web based time/progress updating capability of MS Project
2003? Can progress/time be updated for multiple resources from a single
login, or are multiple logins required. Can multiple users update
time/progress simultaneously?

Many thanks for any help you can provide

 

[John Sitka]

1.Yes, a very large number is allowed
2.Each Project when edited is locked/checked-out single user. Each project can only be calculated
when the complete project is opened in the Desktop application.
ex.
Run thick client->download the whole thing to the desktop->repeat for each project that may have links that need recalculation->do
work.
3.Resources are global but their visability and thus assignment is restricted based on a Resource Breakdown Structure which follows
the corporate hierarchy/functionality/location; whichever is deemed worthy.
4.Web based progress/time reporting is a qeued batch update controlled by the operators of the desktop application.
I have not heard of a solid metric for the functional lag or limit of this data stream reaching the queue and being applied.
Microsoft made the assumption that each resource is a person and thus each resource needs a login, Windows or Project (much like the
SQL Mixed)
and there is no resource reporting agency or control panel option for people who are responsible for multiple resources.


In short
--there is no scheduling/calculation done deyond the desktop other than SQL OLAP reporting
--there is no concept of multiple resource task tracking by a single individual other than the overall builder of the project.
--there is no best fit resource optimization or simulation
--there is seamless integration with Windows Sharepoint but only Windows domain users can get exposed to that.

 

[Mike Glen]

Hi Patrick,

Try posting on the server newsgroup. Please see FAQ Item: 24. Project
Newsgroups. FAQs, companion products and other useful Project information
can be seen at this web address: http://project.mvps.org/faqs.htm

Mike Glen
Project MVP

 

[Patrick]

Thanks for you help John.

1.Yes, a very large number is allowed
2.Each Project when edited is locked/checked-out single user. Each project can only be calculated
when the complete project is opened in the Desktop application.
ex.
Run thick client->download the whole thing to the desktop->repeat for each project that may have links that need recalculation->do
work.
3.Resources are global but their visability and thus assignment is restricted based on a Resource Breakdown Structure which follows
the corporate hierarchy/functionality/location; whichever is deemed worthy.
4.Web based progress/time reporting is a qeued batch update controlled by the operators of the desktop application.
I have not heard of a solid metric for the functional lag or limit of this data stream reaching the queue and being applied.
Microsoft made the assumption that each resource is a person and thus each resource needs a login, Windows or Project (much like the
SQL Mixed)
and there is no resource reporting agency or control panel option for people who are responsible for multiple resources.


In short
--there is no scheduling/calculation done deyond the desktop other than SQL OLAP reporting
--there is no concept of multiple resource task tracking by a single individual other than the overall builder of the project.
--there is no best fit resource optimization or simulation
--there is seamless integration with Windows Sharepoint but only Windows domain users can get exposed to that.

 

[John Sitka]

I think you would miss the multi resource reporting by single indiviual the most.
I only have a vague concept of a chemical plant "turnaround" but I can imagine
I'd want to capture the "at hand" daily best estimates of those who have routinely
been at the dance. Other packages offer the ability to multiproject/reschedule, but
many still need exclsive use of the data or Objects when doing the calculations.


There are two work arounds though,
1.) create the resource accounts under Project Server authentication (NOT windows domain) and have your reporting individual
login/logout multiple times.
(This is actually cheaper too because PWA CALS are per user (Volume licence agreement definition of a user) not per resource or
resource account.)
2.) use resource management projects that have NO logical task linking, only order (sequence tasks) along runs of singular unique
resources.

to expand on number 2.)
The idea is to build Control Plans with no resources only resource reporting directives (codes etc to use as filters to get the
tasks to the right people).
The resource managers then utilize their Projects as control panels only for the resources they are responsible for.

Control Plan A
-Task1
|--Task2
|---Task3
|
| Control Plan B
| -Task1
| --Task2
| ---Task3
|
|Resource Control Panel Project
|-Control Plan S Task100 resource JOE
|--Control Plan T Task5 resource JOE
|---Control Plan A Task2 resource JOE

Control Plan A/Task2
is constrained by an interproject link to
Resource Control Panel Project/Control Plan T Task2

"Control Plan A/Task2" can't start until JOE finishes "Resource Control Panel Project/Control Plan A Task 2"

The constraints on Control Plan A's tasks (external links) turn out to be the resources that are most loaded and causing the most
delay
Once you get a few control points in a few control plans it becomes pretty obvious where to spend your energies to make gains by
identifying
the present and future bottlenecks.

The Resource Control Panel Projects are also a way to collect data in non overwhelming batches from
large resource sets that report frequently.

Really it's just a smoke and mirrors trick to abstract the reporting sets to a size that has some value
when dealing with finite resources and impart simple discipline to the whole system.
It makes very frequent distributed status updates manageable.
It also keeps hero's in check because the impact of confiscating resources
even for brief time spans can be seen before a weeks worth of damage is done
even without a sophisticated optimization. (More like the farmer is out in the back forty. If you
want to tell him to go do something else then get on your hourse ride out there and tell him yourself.
That is a great dose of common sense and flow maintainer when it comes to multi-project contentions.)

A control point showing a dramatic delay in the critical path is an opportunity for gains,
if it had remained hidden, confusing, blurred or most commonly; lied about. It costs money.