2   Managing  Distributed  Data  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
DISCLAIMER STATEMENT

© 2015 - World Food Programme

Information presented in this training document may be considered public information and may be reproduced, distributed or
copied freely, exclusively for non-commercial purposes and unless identified as being subject to third parties’ copyright
protection. World Food Programme (“WFP”) should be cited as the source of the information and any photo credits should be
similarly credited to the author or source, as appropriate. If a third party copyright is indicated on a photo, graphic, or on any
other material, permission to copy these materials must be obtained from the original source.
WFP declines all responsibility for errors or deficiencies in this training document or in the documentation accompanying it,
for the update of the data as well as for any damage that may arise from them. Information in the GIS Training Package is
provided without warranty of any kind, either express or implied, including, without limitation, warranties as to fitness for a
specific purpose, non-infringement, accuracy or completeness. Under no circumstances shall WFP be liable for any direct or
consequential loss, personal injury, property damage, or expense of any nature directly or indirectly incurred or suffered by
any person that is claimed to have resulted from the use of this training document or information included therein.
This training document contains advice, opinions, and statements of several individuals and information providers. WFP does
not necessarily share, represent or endorse any advice, opinion, statement or other information provided by any individual or
information provider. Reliance upon any such advice, opinion, statement, or other information shall also be at the user’s own
risk. WFP shall not be liable to any person for any inaccuracy, error, omission, deletion, defect, failure, alteration or use of
any content herein, regardless of cause, nor for any damages resulting therefrom.
This training document also contains links to third-party websites. These links are provided for the user’s convenience only.
WFP is not responsible for the contents of any linked website and do not control or guarantee the accuracy, relevance,
timeliness, or completeness of any linked information. Further, the inclusion of links is not intended to assign importance to
those sites and the information contained therein, nor is it intended to endorse or recommend any views expressed,
commercial products or services offered on these external sites that are not controlled by WFP. Reference in the GIS Training
Package to any specific commercial products, processes, or services, or the use of any trade, firm or corporation name is
only for the information and convenience of the user and does not constitute an endorsement, recommendation, or judgment
by WFP.
WFP is not liable for any infringements of third party rights as a result of any form of publication by the users of this training
document, or sections thereof, and they alone will be liable.
The designations employed and the presentation of material in maps contained in this document do not imply the expression
of any opinion on the part of WFP concerning the legal or constitutional status of any country, territory, city or sea , or
concerning the delimitation of its frontiers or boundaries.
Nothing related to this statement or this training document may be construed as a waiver, express or implied, by WFP, the
United Nations, and the Food and Agricultural Organization of the United Nations of the privileges and immunities enjoyed by
them pursuant to the 1946 Convention on the Privileges and Immunities of the United Nations, the 1947 Convention on the
Privileges and Immunities of the Specialized Agencies or otherwise under any international or national law, convention or
agreement.
All queries on rights and licenses should be addressed to the Geospatial Support Unit, World Food Programme,

Via C.G. Viola, 68-70, Rome 00148, Italy; e-mail: [email protected].

 
 
Managing  Distributed  Data   3  
 
 
1   Replicas and GeoDatabases .................................................................. 6  
1.1   Replication  types  ..............................................................................................................................................................................  7  
1.2   Data  sharing  in  connected  or  disconnected  environments  ............................................................................................  8  
1.3   Getting  ready  for  replication  .......................................................................................................................................................  9  
1.3.1   Add  GlobalID  column  ................................................................................................................................................................  10  
1.3.2   Registering  data  as  versioned  ...............................................................................................................................................  11  
1.3.3   Define  the  data  to  be  replicated  ...........................................................................................................................................  12  
1.3.4   Replication  and  Metadata  ......................................................................................................................................................  13  
1.3.5   Replication  and  Raster  Data  ..................................................................................................................................................  13  

2   Creating Replicas ................................................................................ 14  

2.1   Creating  replicas  in  a  connected  environment  .................................................................................................................  15  
2.1.1   Creating  a  one-­‐way  or  two-­‐way  replica  ...........................................................................................................................  15  
2.2   Creating  replicas  in  a  disconnected  environment  ...........................................................................................................  23  
2.2.1   Replica  creation  in  the  source  geodatabase  ...................................................................................................................  24  
2.2.2   Replica  transfer  to  the  destination  geodatabase  ..........................................................................................................  27  
2.2.3   Replica  creation  in  the  destination  geodatabase  .........................................................................................................  27  

3   Data Synchronization .......................................................................... 30  

3.1   Synchronizing  Replicas  in  a  connected  environment  ....................................................................................................  31  
3.2   Synchronizing  Replicas  in  a  disconnected  environment  .............................................................................................  35  
3.2.1   Export  a  data  change  message  .............................................................................................................................................  37  
3.2.2   Import  a  Data  Change  Message  ...........................................................................................................................................  39  
3.2.3   Export  an  Acknowledge  Message  ........................................................................................................................................  41  
3.2.4   Import  an  Acknowledge  message  ........................................................................................................................................  42  

4   Managing changes in the layers structure ........................................... 46  

4.1   Applying  schema  changes  in  connected  environments  ................................................................................................  47  
4.1.1   Compare  replica  schema  .........................................................................................................................................................  47  
4.1.2   Import  replica  schema  .............................................................................................................................................................  49  
4.2   Applying  schema  changes  in  disconnected  environments  ..........................................................................................  50  
4.2.1   Export  replica  schema  ..............................................................................................................................................................  51  
4.2.2   Compare  replica  schema  .........................................................................................................................................................  52  
4.2.3   Import  replica  schema  .............................................................................................................................................................  53  
4   Managing  Distributed  Data  
 
 
5   Replica Management ........................................................................... 55  
5.1   Turning  a  disconnected  replica  in  a  connected  one  .......................................................................................................  60  

 
 
Managing  Distributed  Data   5  
 
 
Introduction

 
The GIS training “Managing distributed data” is part of the collection of trainings on
geodatabase management produced by the Geospatial Support Unit (GSU), at the
Emergency Preparedness and Response branch, to support the implementation of
the GIS Infrastructure in the World Food Programme.
This specific training module has been designed with the aim of guiding the user
through the usage of a set of tools provided by ESRI to exchange information
between GeoDatabases distributed in different locations. In particular we will use
replica and synchronization tools described in this training module to share GIS
datasets between WFP offices where GeoDatabases are in place.
Building a network of distributed geodatabase at HQ, Regional Bureau and Country
Office level has been identified as a solution to improve data availability and
performance by alleviating server contention and slow network access to a central
server. Each GIS officer in WFP can work on data stored in a Geodatabase located
in his office and get data produced by others or share his data with others by
taking advantage of Geodatabase replication and synchronization techniques.

In this training module we’ll describe different types of replication (Checkout\Check

in , One-way, Two-way), different options available to create replica workflows in
connected and disconnected environments, how to synchronize data in different
scenarios and manage existing replica.
6   Managing  Distributed  Data  
 
 
Software needed

 
All the tools described in this training module allow replicating and synchronizing
data across ArcSDE GeoDatabases; therefor you must have access to a Desktop,
Workgroup or Enterprise GeoDatabase to go through this training.
In addition such tools are provided by ESRI in a toolbar in ArcMap, so make sure
you have ArcGIS Desktop 10.2.2 Advanced installed in your machine.
In case one of these constraints is not respected, please get in touch with your
regional GIS officer or with the GSU unit at the HQ, by sending an email to
[email protected].

1 Replicas and GeoDatabases

 
 
Geodatabase replication allows you to create copies of data across two or more
geodatabases such that changes to the data may be synchronized.
Replica creation involves the user defining the data to replicate from a source
geodatabase, and then running a process to create replicas. The process copies
data from the source geodatabase to a target geodatabase and creates a replica in
each geodatabase.
The replica describes what data has been replicated and contains the information
needed to synchronize changes. The replica in the source geodatabase is called
parent replica, and the replica in the target geodatabase is the child replica.
Managing  Distributed  Data   7  
 
 
Each ArcSDE geodatabase can host multiple parent and child replica in order to
allow splitting data to replicate in more replicas and run synchronizations with
different frequency and schedule according to frequency to data editing.

1.1 Replication types

 
Three different types of geodatabase replication are available: check out/check in,
one-way replication and two-way replication. According to your specific needs you
can choose one or the other when you create the replica.

Checkout/Check-in replication allows you to replicate a set of data from a parent

geodatabase to a child geodatabase, edit layers in the child and synchronize edits
back to the parent geodatabase when edit session is complete. The main constraint
with this type of replica is that once the data has been synchronized, you can no
longer synchronize additional edits. If additional edits are required, you must
create a new checkout replica.
 
 
One-way replication allows you to replicate a set of data from a parent
geodatabase to a child geodatabase and synchronize data changes multiple times
in a single direction, either from parent to child or from child to parent.
If you create a one-way replica from parent to child you can edit data only in the
parent, the data in the child is considered read only. In case edits are made in the
child, they are overwritten if they conflict with edits applied during synchronization.
One-way replication child to parent works in a similar manner, but in the opposite
way.
One-way replicas persist after synchronization, allowing you to continue sending
data changes.
 
 
8   Managing  Distributed  Data  
 
 
Two-way replication allows you to establish a bidirectional link on a set of data in
a geodatabase in order to send data multiple times between two geodatabases,
from the parent replica to the child replica and from child to parent.
If the same feature is edited in both replica geodatabases, it is detected as a
conflict at the time of synchronization and specific policies can be established to
decide how to resolve conflicts. As for one-way replicas, two-way replicas continue
to exist after synchronization.

Specific workflows have been identified in WFP to exchange GIS data between
offices, taking advantage of the different types of replication provided by ESRI.
In particular for all data produced and maintained at the WFP Headquarters, one-
way replication will be put in place to share it with Regional Bureaux and Country
offices. The same will happen, but opposite way, for all those data produced at
Country Office Level, while two-ways replication will be used to exchange data that
could be edited by more GIS officers at the HQ, RBs or COs, such as logistics
assets, WFP facilities or warehouses locations.

1.2 Data sharing in connected or disconnected environments

Data replication and synchronization tools provided by ESRI have been designed to
work in different operating contexts; in scenarios where connectivity between
offices is strong and reliable, ensuring that replica geodatabases are accessible on
the network at all times, and environments where internet connectivity is weak or
not available.

When planning which replication type or synchronization mechanism to use, you

must assess the amount of resources available in terms of network bandwidth and
latency between offices to choose of creating replicas with procedures thought for
Managing  Distributed  Data   9  
 
 
connected environments and then synchronize changes on-line or put in place
“disconnected procedures”.
In general the WFP’s network cannot be considered very reliable and in many
offices Internet capacity is very weak. For this reasons the data sharing strategy
must be designed carefully to ensure data availability at all times.
In the second chapter we will describe all options available to create replicas and
then a specific workflow, which makes use of a combination of various techniques
to create replica without a strong use of Internet connectivity and then synchronize
data on-line when internet connectivity is available and use “disconnected”
workflows when geodatabases in different locations are not connected.

1.3 Getting ready for replication

Before creating the replica you must determine which datasets to replicate and
prepare them ensuring they meet the following requirements:

o The database user in the connection file used to create the replica must have
write aces to the data.

o Each dataset must have a GlobalID column.

o All data must be registered as fully versioned; It cannot be versioned with

the option to move edits to base.

o Spatial data must be stored in high-precision spatial reference.

Most of the layers you have should be already stored in a high precision
spatial reference. In case you get an error related to spatial reference
10   Managing  Distributed  Data  
 
 
precision during the creation of a replica you can migrate your data to high
precision follow instructions available on the ESRI portal:
http://resources.arcgis.com/en/help/main/10.2/index.html#/Migrating_to_hi
gh_precision/003n00000033000000/

1.3.1 Add GlobalID column

 
You can add global IDs to a dataset in a geodatabase by right-clicking it in the
Catalog tree and clicking Add Global IDs as reported in the figure below.

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
This command is available for stand-alone feature classes, feature datasets, tables
and attributed relationship classes in a geodatabase but it cannot be executed on
individual datasets in a feature dataset. If you add a feature class to a feature
datasets that already has GlobalIDs for all feature classes, you have to run again
the Add GlobalID command on the feature dataset. It will add the GlobalID only to
the feature classes without GlobalID columns.
 
 
 
Managing  Distributed  Data   11  
 
 
1.3.2 Registering data as versioned
 
You can only register your data as versioned when it is not in use by other people
or applications, because an exclusive lock is required to ensure the dataset is not
changing while the adds table is being created.
 

To register data as versioned follow steps below:  

 
 
1. Start ArcCatalog or ArcMap and connect to the geodatabase that contains the
feature dataset, feature class, or table you want to version, using a
connection file created with the owner of the object you want to version.
2. In the Catalog tree, right-click the dataset, point to Manage, then
click Register As Versioned.
3. The Register As Versioned dialog box opens.

4. Make sure the checkbox Register the selected objects with the option to
move edits to base IS NOT selected.
5. Click OK.
12   Managing  Distributed  Data  
 
 
 
 

1.3.3 Define the data to be replicated

 
You can limit the amount of data to replicate for each dataset by choosing to
replicate only a subset of rows, a certain extent of your layers or even just the
schema.
You can apply three types of filters on the data you replicate:

Spatial: You can limit the extent included in the replica by using the current view
extent of the ArcMAP document you are using to create the replica. To do
this you have to select a specific option in the Advanced Create Replica
option that will be described in the chapter 2. All the features intersecting
this filter will be included in the replica and will be affected by
synchronization as per image below:

 
Selections: If there is a selection of certain objects in the datasets you are
including in a replica, only these features will be included in the
replica.

QueryDefs: Definition queries can be used to filter the content of feature classes
and tables.
Managing  Distributed  Data   13  
 
 
If more than one filter is used, the intersection of all filters is applied.

1.3.4 Replication and Metadata

Metadata for the data you choose to replicate is copied during the replica creation
process. However, changes to the metadata are not applied during replica
synchronization.

1.3.5 Replication and Raster Data

 
Raster catalogs, raster datasets and mosaic datasets cannot be versioned in ArcGIS
for Desktop and, therefore, cannot be replicated. However, the Create Replica
wizard will extract a raster dataset, raster catalog or mosaic dataset from the
source geodatabase if it is in the Arc Map document. Since the data is extracted,
any edits made to the raster will not be transferred to the relative replica during
synchronization.
The spatial extent of the replication area is used to determine the portion of the
raster to extract. For raster datasets, the data is clipped to the spatial extent. For
raster catalogs and mosaic datasets, the entire raster dataset is extracted for each
raster dataset intersecting the spatial extent.
 

1.3.6 Replication and joined tables

 
 
In case layers are joined with tables in the same geodatabase, when replica are
crated including these layers, joined tables will be replicated as well but the join is
not kept. It means that once the replica creation is finalized, the join between
layers and tables must be recreated in the destination geodatabase.
 
 
14   Managing  Distributed  Data  
 
 
2 Creating Replicas
 
All tasks related to replica creation and management can be executed through
ArcMap using a specific toolbar called Distributed GeoDatabases. You can display it
in ArcMap by selecting the tab Customize, then choose Toolbars and click on
Distribute GeoDatabases.
At this point the toolbar reported below should appear:
 
 
 
 
 
 
This toolbar contains a set of tools that allow you to create a replica, synchronize
changes, apply schema changes, manage existing replica and more. In this section
of the training we’ll explore the first tool for creating replica through a wizard.
It will allow you to define what data should be replicated, what kind of replica you
would like to create, and if you are putting in place a replica in a connected
environment, which is the destination geodatabase.
We will initially describe the process to Create Replicas in Connected Environments
and then additional steps needed in case your Internet connectivity is not reliable
or insufficient.
Before starting please consider that:

o In case you want to add to the replica more layers stored in the same feature
dataset, you can add just one of them to the ArcMap document because at a
later stage the Creating Replica Wizard will present you a list of objects you
want to include in the replica, taking into consideration the entire feature
dataset for each layer in the ArcMap document;
Managing  Distributed  Data   15  
 
 
2.1 Creating replicas in a connected environment

If you have already prepared your data for replication, you can use the Create
Replica wizard in ArcMap to create checkout, one-way or two-way replicas.
For putting in place a replica in a connected environment two connection files are
needed: one connection file to your geodatabase and one connection file to the
destination geodatabase. Both connection files but must connect to the
geodatabase with a user with privileges to create new objects in the respective
geodatabase, otherwise once the replica will be in place you will have limited
privileges (for instance you won’t be able to change the schema of a layer adding
or removing fields).

2.1.1 Creating a one-way or two-way replica

You find below the steps to create a one-way or two-way replica. In the example
we have chosen to include in the replica a layer containing Earthquake events
stored in a feature dataset called NHR.

1. Open a new ArcMap document;

2. Make sure the objects you want to include in the replica have a GlonalID
column and are Registered as Versioned;

3. Connect to your local geodatabase using the connection file and load into the
ArcMap document the objects you want to include in the replica. In this case we
have the layer: wld_nhr_EQepicenters_usgs;
16   Managing  Distributed  Data  
 
 

4. Add any filtering criteria may be required. Spatial, Selections or Definition

Queries. In this case we have zoomed in to include in the map extent only an
area around Africa.
Managing  Distributed  Data   17  
 
 
5. If the Distributed Geodatabase toolbar is not already open, click Customize on
the main menu, point to Toolbars, and click Distributed Geodatabase.

6. If layers that can be replicated are in your ArcMap document, the Create
Replica button on the Distributed Geodatabase toolbar will be active, click it
to activate the Create Replica wizard.

7. If your ArcMap document includes data from more than one ArcSDE
geodatabase you are prompted to choose which ArcSDE geodatabase you want
to work with.

8. At this step a the Creating Replica wizard will appear, letting you choose to
create either a check-out, a two-way, or a one-way replica. In this instruction
we are focusing on these last two replica types, for which the process for
creating are very similar. The main difference is that, in case you choose a one-
way replica, you have to select if it must be parent-to-child or child-to-parent.
In this example we select a two-way replica.
18   Managing  Distributed  Data  
 
 

9. On the next panel, choose to replicate data or register existing data only.
The option to register existing data only allows you to create a replica between
two geodatabases containing the same data. It may be useful if your internet
connection is reliable but available bandwidth is limited. In this case you can
send data to the destination database by other means (FTP, physical drives etc)
and then create the replica. In this example we choose to replicate also the
data.

10. You can choose to replicate to a Geodatabase or an XML workspace. To create a

Replica using the “Connected” workflow you must specify the connection file to
connect to the destination geodatabase.
While the XML option is used to create the replica using the procedure for
disconnected environments, which will be described later.
Browse to the connection file to which you want to replicate data.

11. Type the name of the replica you are creating.

Managing  Distributed  Data   19  
 
 
You should select a name that then helps you identify the main information of
the replica, such as source, destination and data included. In this example we
use as replica name
HQ_OMB_Earthquakes because we are creating a replica between the
enterprise geodatabase at the WFP headquarters and a geodatabase at the
regional office in Bangkok including in the replica earthquake epicenters.

12. Check the Show advanced options box and click next to proceed.

13. The first panel of the Advanced Create Replica Options offers the choice
between a Full or Simple model. Keep the default Full model and click next.

14. The next window let you decide if you want to limit the replica to a specific
extent (the one in the ArcMap document window selected at the beginning),
use the full extent of the data or specify manually an extent.
Through this window you can also exclude individual layers or tables from the
replica, unchecking the check box associated with that layer or table.
As mentioned before, if you selected in the first step a layer stored in a
Feature Dataset, a list of all the layers in that FD will be presented.
To include in the replica only your Earthquake layer, uncheck all the others.

You can also change the dataset's name when it is replicated, under the Target
Name column. It may be useful if you are replicating only a portion of the data.

If you have previously chosen the option to register existing data only, then
each name in the Target Name column will offer a drop-down list of available
datasets in the target geodatabase to choose from.
20   Managing  Distributed  Data  
 
 

Each entry in the Check Out column is a combo box of options. The options
always include All Features and Schema Only. If a particular layer or table
has a selection set or definition query defined, the options may also
include Selected Features Only, All Features in Def. Query, and Selected
Features in Def. Query.
Select All Features and uncheck the Use Spatial Extent box if you do not want to
apply any data filters.
Uncheck the Replicate related data check box if you do not want to replicate
any related data and click Next. This option may be useful if relationship classes
are associated to the layers in the replica.

o When replicating tables the default option is to replicate the schema only; in
case you want replicate the data as well, make sure to select All Records in
the Check Out field, as shown in the image below;
Managing  Distributed  Data   21  
 
 

15. For two-way replication, there is an additional panel in the Advanced Create
Replica Options dialog box that lets you choose whether the parent or child
will be the initial data sender. This option is only important in disconnected
systems.

16. Click Next in the following window.

22   Managing  Distributed  Data  
 
 

17. At this step you can choose what you want to do once the replication has been
completed
Take no further action — This is the default option. If you choose this
option, the replica will be created with the ArcMap document showing the
original data.
Change the layers and tables to point to the replicated data — The
current ArcMap document will be modified to point to the data in the
replica geodatabase, preserving all symbology.
Save a copy of this map document with the layers and tables
pointing at the replicated data — A new ArcMap document referencing
the data in the replica geodatabase with symbology preserved will be
created.

18. Click Summary to review the parameters for the current replica.

19. Click Finish to start replicating the data. The status of the replication
operation will be monitored in a progress dialog box.
Managing  Distributed  Data   23  
 
 
If you are interested in getting to know how to create a checkout replica please
follow the guide on the ESRI portal:

http://resources.arcgis.com/en/help/main/10.2/index.html#/Creating_a_check_out
_replica/003n000000v5000000/

2.2 Creating replicas in a disconnected environment

In this section we’ll describe the process to create a replica when Internet
connectivity is not reliable or weak. After several tests at Country Office and
Regional Bureau level we can say that is the process that has more chances to
succeed when putting in place replicas in WFP, considering networking constraints
in our organization.
The process is slightly more involved than in a connected environment simply
because there are more things to remember and steps to follow.
It can be summarized in a 3 phases process:

1) Create the replica in the source geodatabase;

2) Sent it to the person managing the destination geodatabase;
3) Import the replica in the destination geodatabase.

At some point, when we described the steps to put in place a replica in a connected
environment, we choice to send it to the destination geodatabase online by
selecting a connection file. In this case we have to save the replica in an XML file,
because we cannot ensure that the destination geodatabase is reachable remotely.
Once the replica is created, it must be sent to the destination geodatabase by
FTP transfer, email or physical drives and then imported in the destination
geodatabase to finalize the replica creation process.
24   Managing  Distributed  Data  
 
 

2.2.1 Replica creation in the source geodatabase

We’ll now describe the step-by-step process to create a one-way replica in a

disconnected environment, simulating a scenario where the source geodatabase
(Parent) is at the WFP Country Office in Myanmar, the destination geodatabase
(Child) at the WFP HQ in Rome and including in the replica a layer containing WFP
facilities.

1. Open a new ArcMap document;

2. Make sure the layer you want to include in the replica has a GlobalID column
and is Registered as Versioned;

3. Connect to your local geodatabase using the connection file and load into the
ArcMap document the layer you want to include in the replica. In this case we
have the layer: mmr_poi_facilities_wfp
Managing  Distributed  Data   25  
 
 

4. Add any filtering criteria may be required. Spatial, Selections or Definition

Queries. In this case we do not add any filtering criteria.

5. Click the Create Replica button on the Distributed Geodatabase toolbar to

activate the Create Replica wizard.

6. Choose to create a one-way replica Parent to child.

7. On the next panel

• Choose to replicate also the data by checking the box Data;
• Choose to create the replica into an XML document by browsing the
folder where you want to save it and choose a name for your replica.
You should select a name that then helps you identify the main
information of the replica, such as source, destination and data
included. In this example we use:
Replica_Myanmar_to_HQ_WFPFacilities.xml ;
• Type the name of the replica you are creating. In this example we use
as replica name: Myanmar_to_HQ_WFPFacilities ;
• Check the Show advanced options box and click next to proceed.
26   Managing  Distributed  Data  
 
 
8. In the first panel of the Advanced Create Replica Options keep the default
Full model and click next.

9. In the next window choose to replicate the full extent of the data.
In the list of items to checkout you will notice that there is not only the WFP
facilities layer but also a layer containing warehouses locations. This happens
because both layers are stored in the same feature dataset (WFP).
In this example we uncheck the warehouses layer to include in the replica only
WFP facilities.

10. Click Next

Managing  Distributed  Data   27  
 
 
11. Choose to Take no further action and click Finish to finalize the replica
creation process in the source geodatabase.

Once the process is completed the replica is registered in the source geodatabase
and all the data needed to create the replica in the destination geodatabase is
stored in the xml file.

2.2.2 Replica transfer to the destination geodatabase

 
At this step you must send the xml file produced in the previous phase to the
person managing the destination geodatabase to complete the replica creation
process from your side.

Please note that in most of the cases the generated XML will be too big to be sent
by email. In this case the preferred option to send the file is the FTP transfer. FTP
servers have been made available at HQ and at each Regional Bureau to support
the implementation of the GIS infrastructure in WFP. Therefore you should get the
necessary information to access both FTP servers (HQ and at your reference RB),
before creating this kind of replica.

2.2.3 Replica creation in the destination geodatabase

 
Once the person managing the destination geodatabase receives the XML
document containing the replica, the final phase of the replica creation process can
be performed by importing all the data and registering the replica in the destination
geodatabase.
28   Managing  Distributed  Data  
 
 
Both results can be achieved through one single operation, by importing the XML
document into the destination geodatabase ad an XML workspace.

We report here the steps needed:

 
1. In the Catalog tree, right-click the geodatabase into which you want to import
the replica and point to Import > XML Workspace Document.
The window shown in the image below will appear.

2. Choose whether to import the geodatabase and its data by clicking the Data
checkbox.
3. Click the open folder button and navigate to the XML document containing the
replica, select it and click Open.
4. Click Next.
5. The window shown in the image below will appear. It presents a list of objects
that will be imported in the destination geodatabase, any related data (Domain,
Relationship Classes, Topologies etc.) will also appear.
Managing  Distributed  Data   29  
 
 
Any naming conflicts are displayed in red. To change a suggested name in
the Target Name column, overwrite the name.

6. Click Finish to complete the process.

Additional options are available to create replicas in both connected and

disconnected environments, when for instance the same data is already available in
the source and destination geodatabases or you want to replicate only the schema
of your data.
If you are interested in such procedures please refers to the documentation on the
ESRI platform, as they are out of the scope of this training document.
 
http://resources.arcgis.com/en/help/main/10.2/index.html#/Replicas_created_with_the_option_t
o_register_with_existing_data/003n000000vq000000/  
 
http://resources.arcgis.com/en/help/main/10.2/index.html#/Reusing_a_schema_to_create_a_repl
ica/003n000000w3000000/  
30   Managing  Distributed  Data  
 
 
3 Data Synchronization

Once a replica between two geodatabases is in place data changes can be

exchanged between them using synchronization procedures.
Data changes include inserts, updates and delete of features in existing layers or
tables but not changes in the schema structure (add or remove fields or layers
etc.). These must be managed with a specific workflow called Apply Schema
changes that will be described later.
Data synchronization is performed through a specific tool in the Distributed
Geodatabases toolbar, after connecting to the database originator or the changes
with the same database user as that used to create the replica, or the geodatabase
administrator.

For two-way and one-way replication, the same filters and relationship class rules
used in replica creation are applied during synchronization, with the exception of
filters based on a selection set. When determining the changes to send, all edits in
each replica dataset that have been applied since the last synchronization are
evaluated. If an edit satisfies the replica's filters, it will be synchronized.
 
Additionally the geodatabase synchronization system detects errors occurring
during synchronizations and rollbacks in case needed. Any changes that have been
applied are removed, and the system is put back as it was before the
synchronization.

Synchronization is supported in both a connected and a disconnected environment.

In a connected environment, message exchange is handled by the system, while in
a disconnected environment; you need to manage the message exchange. Both
processes are fully described in this chapter.
 
Managing  Distributed  Data   31  
 
 
If you are using advanced geodatabase features, such as geometric networks,
topologies or versioning, it’s strongly suggested to have a look at the specific
documentation on the ESRI portal.
 
http://resources.arcgis.com/en/help/main/10.2/index.html#/Synchronization_and_versioning/0
03n000000vt000000/  
 
 

3.1 Synchronizing Replicas in a connected environment

 
 
Synchronize connected replicas is an easy procedure that can be carried out using
the Synchronization Changes wizard in ArcMap.

You can connect to your geodatabase and choose the replica to synchronize. All the
required message exchanges needed to complete synchronization are executed by
the system. You never have to be concerned with message exchange or which
replica is the sender or receiver.
Moreover connected synchronization allows you to choose the direction in which the
changes will be sent in a two-way replica. For example, you can send changes from
the parent replica to the child replica or from the child replica to the parent replica,
or in both directions. If you choose both directions, changes are first sent in one
direction and then sent in the opposite direction, all in one operation.

Here the step-by-step process:

 
1. Edit the layer included in your replica;

2. Open the Synchronize Wizard by clicking the Synchronize button on the

Distributed Geodatabase toolbar;
 
 
 
 
32   Managing  Distributed  Data  
 
 
 
3. On the first panel of the Synchronize wizard, choose the replica that you want to
synchronize. In this example we want to synchronize the two-way replica
HQ_to_OMB_Earthquakes
Once selected, the connection information for geodatabase 2 is automatically
filled in.

4. For two-way replicas, choose the direction in which you would like to send
changes. For checkout replicas, the only option available is to send changes
from the child replica to parent replica. For one-way replicas, the only option
available is to send changes from the parent replica to the child replica, or if it's
a child-to-parent replica.
Managing  Distributed  Data   33  
 
 
5. Click Next.

6. For checkout replicas, there is an option to reconcile and post with the parent
version upon synchronization. For two-way and one-way replicas, this is always
checked on.

7. Next, choose to define conflicts By Object or By Attribute.

8. Defining conflicts by object detects conflicts by row, while defining them by

attribute detects conflicts by column.

9. Choose how you want conflicts resolved. (the first two are usually preferred)

In favor of the geodatabase1 — In this case, the edits in geodatabase1 are

automatically used over the edits in geodatabase2 if there is a conflict. Since the
34   Managing  Distributed  Data  
 
 
conflicts are resolved automatically, the replica is never in a conflict state after
synchronizing with this policy.

In favor of the geodatabase2 — In this case, the edits in geodatabase2 is

automatically used over the edits in geodatabase1 if there is a conflict. Since the
conflicts are resolved automatically, the replica is never in a conflict state after
synchronizing with this policy.

Manual — With this policy, if a conflict occurs, the reconcile operation is aborted
and the replica is marked as in conflict. This gives you an opportunity to perform
the reconcile afterwards either manually or by running some custom reconcile
code. Once the reconcile is applied and the changes posted to the replica
version, the replica is no longer in conflict. While the replica is in conflict it can
continue to receive changes but cannot send changes.

10. Click Finish.

 
Managing  Distributed  Data   35  
 
 
3.2 Synchronizing Replicas in a disconnected environment

In case you need to synchronize a “disconnected” replica, you have to go through a

process of manually exchanging messages between replicas in the source and
destination geodatabases. This is needed because the geodatabase that sends
changes must be aware if the synchronization on the other hand completed
successfully or changes need to be synchronized again.
For this reason two types of messages must be exchanged: data change messages
and acknowledgement messages.

This happens through the multi-steps process described in the image below:

The data sender exports changes to be applied to the relative replica into a Data
Change Message and send it to the receiver. The data receiver imports this
message into his geodatabase and exports an Acknowledgment Message and
send it to the data sender to confirm that data changes were received and
processed. The data sender imports the Acknowledge Message into its geodatabase
to complete the process.
36   Managing  Distributed  Data  
 
 
It is important for the data receiver to export acknowledgment messages as often
as possible. If no acknowledgment messages are received, the data sender resends
changes by default, and maintains the information needed to resend changes until
those changes are acknowledged. As a result, the data sender's geodatabase can
become large, and subsequent data change messages can also become large.

If you have particular needs that are not covered by this workflow please refer to
the ESRI’s documentation or get in touch with your GIS focal point in WFP.

http://resources.arcgis.com/en/help/main/10.2/index.html#/Disconnected_synchronizatio
n/003n000000v2000000/
We’ll now describe the entire process with a practical example, using the
disconnected replica created in the second Chapter
(Myanmar_to_HQ_WFPFacilities)

After adding a new feature in the source geodatabase we start the process to send
this change to the destination geodatabase through a disconnected
synchronization.
Managing  Distributed  Data   37  
 
 
3.2.1 Export a data change message

1. Add the layer to an ArcMap document; (Highlighted in a red box the new point)

2. Open the Distributed Geodatabase toolbar and click the Export Data Changes
Message button.

3. In the wizard, select the replica from which you would like to export data
changes and specify if you want to save the message in a delta database or an
XML file. We suggest using an XML file if there is few data to synchronize, while
the geodatabase option is preferred in case you have done a lot of edits.
38   Managing  Distributed  Data  
 
 

Specify a self-explanatory name for the Data Change message. It should contain at
least the name of the Replica and the date as per the example below:

ExportaDataChanges_Myanmar_To_HQ_BaseLayers_17122014.xml

4. The last 3 check boxes are useful to manage particular needs. In general you
should leave the last 2 checked.

For reference we report some info about these checkboxes.

By combining the last two check boxes, you have three additional potential options:

1) Include all unacknowledged data changes and new data changes since the
last export.
2) Include only those unacknowledged data changes that you have not received
an acknowledgment message for.
Managing  Distributed  Data   39  
 
 

3) Don't include any data change messages. This option is useful for two-way
replica, for sending a message to switch roles without sending any data.

Once the Data Change Message is saved in you file system, you have to send it to
the person managing the destination geodatabase as described in the section
Creating replicas in a disconnected environment.

3.2.2 Import a Data Change Message

In order to synchronize edits made in the source geodatabase the person managing
the destination geodatabase must import the Data Change Message received.

1. Add one of the layers included in the replica that needs to be synchronized to
an ArcMap document and click the Import Message button that can be found
in the Distributed Geodatabase toolbar.

2. If prompted, choose the replica geodatabase you want to import a message to.
It is needed only if in the ArcMap document there are layers stored in different
geodatabases.

3. Choose the delta file you would like to import. (XML or Geodatabase)

4. In case of a two-ways replica, choose between column- and row-level conflict

detection and how to resolve conflicts:
40   Managing  Distributed  Data  
 
 
• In favor of the database — In this case, the edits of the database
importing the changes are used over the edits in the delta file is there is a
conflict. Since the conflicts are resolved automatically, the replica is never
in a conflict state after importing with this policy.
• In favor of the imported changes — Here, the edits defined in the delta
file are used over the edits of the database importing the changes if there is
a conflict. Since the conflicts are resolved automatically, the replica is never
in a conflict state after importing with this policy.
• Manually at a later time—With this policy, if a conflict occurs the
reconcile operation is aborted and the replica is marked as in conflict. This
gives you an opportunity to perform the reconcile afterwards, either
manually or by running some custom reconcile code. Once the reconcile is
applied and the changes posted to the replica version, the replica is no
longer in conflict. While the replica is in conflict, it can continue to receive
changes but cannot send changes.

5. Click Finish to finalize the process. Now the new point is present also in the
destination geodatabase.
Managing  Distributed  Data   41  
 
 
3.2.3 Export an Acknowledge Message

Once changes are imported into the destination geodatabase, an Acknowledge

message must be produced and sent to the person managing the source
geodatabase to confirm that the process has been completed successfully.

1. Ensure one of the layers included in the replica that needs to be

acknowledged is in the ArcMap document and click the Export
Acknowledge Message button that can be found in the Distributed
Geodatabase toolbar.

2. If prompted, choose the replica geodatabase you want to import a message

to. It is needed only if in the ArcMap document there are layers stored in
different geodatabases.

3. In the wizard, select the replica from which you would like to export the
acknowledge message and specify its name.
42   Managing  Distributed  Data  
 
 

Specify a self-explanatory name. It should contain at least the name of the Replica and
the date as per the example below:

AcknowledgeMessage_Myanmar_To_HQ_Earthquakes_20012015.xml

4. Click Finish

Once the Acknowledge Message is saved in your file system, you have to send it to
the person managing the source geodatabase as described in the section Creating
replicas in a disconnected environment.

3.2.4 Import an Acknowledge message

This last phase finalizes the process of synchronized data in a disconnected

environment. The person managing the source geodatabase must import the
received Acknowledge Message by following the steps below.

1. Add one of the layers included in the replica that needs to be synchronized to
an ArcMap document and click the Import Message button that can be
found in the Distributed Geodatabase toolbar.

2. If prompted, choose the replica geodatabase you want to import a message

to.
Managing  Distributed  Data   43  
 
 
It is needed only if in the ArcMap document there are layers stored in different
geodatabases.

3. Choose the Acknowledge Message you would like to import.

4. Click Finish to finalize the process.

44   Managing  Distributed  Data  
 
 
3.2.5 Synchronize “connected” replicas automatically

In a system where there are many replicas and/or frequent synchronizations,

guarantee that data is efficiently synchronized can be very time and bandwidth
consuming.
For support these use cases, it is possible to create replication agents to automate
the synchronization process. Replication agents work by automatically connecting
to replicated geodatabases and performing synchronizations. In this case, end
users do not have to explicitly synchronize their databases, as synchronization
happens automatically.
In a connected environment, replication agents can be built using geoprocessing
tools, export them to Python scripts and then execute them using scheduling
software, such as the Windows scheduler, so that they can be run on a regular
basis. For example, you may want to schedule a synchronization between two
enterprise geodatabases once a week at a nonpeak time.

Below we report an example of script produced to synchronize a replica containing

natural hazard datasets from the geodatabase 1 at the WFP HQ and the
geodatabase 2 at the regional office in Bangkok (OMB).
You need to provide to the script:

• The connection file to connect to the Geodatabase 1 (wfp_hq_esri_HQ.sde)

• The connection file to connect to the Geodatabase 2
(hq_hq_esri_OMB.sde)
• Replica name ( HQ_to_OMB_NaturalHazards)
Managing  Distributed  Data   45  
 
 
from arcpy import env
import sys
print(sys.version)
print("arcpy imported")

# Set workspace
workspace = r"C:\wfp_sdi_confs\synchro_script"

# Set local variables

replica_gdb1 = r"C:\wfp_sdi_confs\synchro_script\wfp__hq_esri_HQ.sde"
print("defined geodb 1")
replica_gdb2 = r"C:\wfp_sdi_confs\synchro_script\hq__hq_esri_OMB.sde"
replica_name = "wfp.HQ_to_OMB_NaturalHazards"
sync_direction = "FROM_GEODATABASE1_TO_2"
conflict_policy = "IN_FAVOR_OF_GDB1"
# Not applicable for one way replicas, there is not conflict detection.
conflict_detection = "BY_OBJECT"
# Not applicable for one way replicas, there is not conflict detection.
reconcile = ""
# Only applicable for Checkout replicas

print("Synchronization starting at " + str(datetime.datetime.now()))

print("#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-#-")
print("Please don't close this window, wait for next message.")
# Execute SynchronizeChanges.
arcpy.SynchronizeChanges_management(replica_gdb1, replica_name, replica_gdb2,
sync_direction, conflict_policy,
conflict_detection, reconcile)

print("Synchronization ended at " + str(datetime.datetime.now()))

del arcpy
print("Now safe to close. Press ENTER.")
# raw_input()
46   Managing  Distributed  Data  
 
 
4 Managing changes in the layers structure

When a replica is created, data and schema are copied from the source
geodatabase (Parent) to the destination geodatabase (Child).
The data includes the rows to be replicated from the datasets in the replica.
The schema consists of the fields, domains, subtypes, and other properties that
describe the replicated data.
Initially, the schemas are identical on both replicas, but over time, changes might
be applied to the schema in each geodatabase.

Specific procedures have been designed by ESRI to support replica and

synchronizations when the schema of one or more objects in the replica is changed
(for instance when you add or remove a field). Even though, it is best to avoid
schema changes because they can lead to inconsistencies among replicas, and the
extra task of applying schema changes can increase performance costs. However,
there are cases when schema changes must be applied.

In these cases before synchronizing the replica you have to run through a Schema
Change process. It first compares the schema of layers in both geodatabases
(source and destination) then modifies the structure of the geodatabase receiving
the Schema Change by modifying the structure.

These procedures are slightly different for replicas in connected or disconnected

environments.
Managing  Distributed  Data   47  
 
 
4.1 Applying schema changes in connected environments

In case you are working with a “connected“ replica you can execute a schema
change using the Compare Replica Schemas tool to check differences in the
structure of the layers in the two geodatabases and then the Import Schema
Changes tool to apply changes on the destination geodatabase.

4.1.1 Compare replica schema

In this example we apply a Schema Change after having added a field

(schematest) to the Earthquake layer in the source geodatabase.

1. Connect to the destination geodatabase and add the layers included in the
replica that needs to be synchronized to an ArcMap document and click the
Compare Replica Schema button that can be found in the Distributed
Geodatabase toolbar.
48   Managing  Distributed  Data  
 
 
2. Specify or browse the connection file to the source geodatabase.

3. Use the Replica name combo box to choose the replica.

4. Specify or browse for an output replica schema changes XML file.

Specify a self-explanatory name. It should contain at least the type of

message (SchemaChange in this case), name of the Replica and the date as
per the example below:

SchemaChange_HQ_TO_OMB_Earthquakes_20012015.xml

5. Click Finish.
Managing  Distributed  Data   49  
 
 
4.1.2 Import replica schema

1. Connect to the destination geodatabase and add one of the layers included in
the replica that needs to be synchronized to an ArcMap document and click
the Import Schema Changes button that can be found in the Distributed
Geodatabase toolbar.

2. Browse for the replica schema changes file. Replica name and Replica type
will be automatically filled in.

3. Click Next
50   Managing  Distributed  Data  
 
 

4. The second dialog box lists the differences between the two schemas. Check
the check boxes under the Apply column for the changes you want to apply
to the replica schema.

5. Click Finish.

At this point the Earthquake layer in the destination geodatabase should have the
“schematest” field and any synchronization should work properly.

4.2 Applying schema changes in disconnected environments

In case you are working with a “disconnected“ replica, before comparing the
schemas and import schema changes you must export your schema to a file, send
it to the destination replica where it can be compared and imported.
At this scope we’ll use an additional tool provided in the Distribution Geodatabage
toolbar called Export Replica Schema.
Managing  Distributed  Data   51  
 
 
For this example we’ll remove the int_staff field from the WFPFacilities layer
which we included in a disconnected replica in the Chapter 2

4.2.1 Export replica schema

1. Connect to the source geodatabase, add the layer included in the replica that
needs to be synchronized to an ArcMap document and click the Export
Replica Schema button that can be found in the Distributed Geodatabase
toolbar.

1. Specify a name for the XML file that is to be created.

2. Choose the replica to export.

3. Click Finish.
52   Managing  Distributed  Data  
 
 
At this point you have to send this export to the person managing the destination
database who will compare the schema and import changes.

4.2.2 Compare replica schema

1. Connect to the destination geodatabase and add the layers included in the
replica that needs to be synchronized to an ArcMap document and click the
Compare Replica Schema button.

2. Browse the replica schema file received.

3. Use the Replica name combo box to choose the replica.
4. Specify an output replica schema changes XML file.

Specify a self-explanatory name. It should contain at least the type of

message (OutputSchemaChange in this case), name of the Replica and the
date as per the example below:

OutputSchemaChange_Myanmar_to_HQ_20012015.xml
Managing  Distributed  Data   53  
 
 
5. Click Finish.

4.2.3 Import replica schema

 

1. Connect to the destination geodatabase and add one of the layers included in
the replica that needs to be synchronized to an ArcMap document and click
the Import Schema Changes button.

2. Browse for the replica schema changes file. Replica name and Replica type
will be automatically filled in.

3. Click Next

4. The second dialog box lists the differences between the two schemas. Check
the check boxes under the Apply column for the changes you want to apply
to the replica schema.
54   Managing  Distributed  Data  
 
 

5. Click Finish.

At this point the Schema Change process in a disconnected environment is finished

and the int_staff field should have been removed from the WFPFacilities layer in
the destination geodatabase.
Managing  Distributed  Data   55  
 
 
5 Replica Management
 
Going through the previous chapters you may have noticed that working with
replicas involves many procedures and things to remember. Is it a checkout, one-
way or two-way replica? Who is child or parent in a replica pair? Can we
synchronize data using the workflow for connected environments?
Most of these questions can be answered using a tool available in the Distributed
Geodatabase toolbar called Replica Manager.

You can use this tool to rename, refresh, and review the properties of each replica
as well as remove datasets from a replica. You may also view all replicas with a
role of parent or child; list only checkout/check-in, one-way, two-way, or all types
of replicas; and view the replica log. This utility is available in both ArcCatalog and
ArcMap.
To open the Replica Manager, click the Manage Replicas button on the Distributed
Geodatabase toolbar.

In the example below replicas created in the Second Chapter are listed.
56   Managing  Distributed  Data  
 
 
For each replica you can access information regarding:

• Name

• Owner — The user who created the replica

• Type — The type of replica created: Check out/Check in, One way, or Two way

• Role — Shows whether the replica is a parent or child

• Status — Displays whether the replica is currently a data sender or a data

receiver

• Conflicts — Displays Yes if there are any outstanding conflicts to be resolved

with this replica and No if there are not

• Version — The replica version set during replica creation

• Date Created — The date and time that the replica was created

By right-clicking an individual replica, you bring up a context menu with the

following options:

• Rename — Renames the replica; type a new name and press ENTER.

• Unregister — Unregisters the replica; if a synchronization version exists, it will

also be removed.

• View log — Opens the Replica Log, which maintains a record for each data
message sent or received by the replica.

• Refresh — Refreshes the replica; the latest state of the replica properties are
displayed.

• Validate schema—Validates the replica schema; a progress bar monitors the

status of the validation. This automatically corrects a replica that contains
Managing  Distributed  Data   57  
 
 
invalid datasets. Datasets can become invalid if they have been renamed,
deleted, or unversioned.

• Properties—Opens the Replica Properties dialog box. Replica properties are

discussed in more detail below.

This last option is particularly interesting because it allows changing some replica
parameters that are fundamental for the workflows identified in WFP.
It contains three tabs: General, Description, and Advanced.

The General tab shows some additional replica parameters, such as the replica
and synchronization versions.
58   Managing  Distributed  Data  
 
 
The Description tab shows the datasets that are included in the replica and allows
to remove them by right clicking and choosing UnRegister from Replica;
however, they still remains in the geodatabase.

You can also see if any filters were applied to the replica dataset during replica
creation by choosing View filters.

The Advanced tab in the Replica Properties dialog box shows information about
the generation numbers associated with the replica. The generation number is a
number maintained by the geodatabase, which keeps track of messages being sent
and received by the replica. For example, the first data change message sent from
one replica to its relative would make the current generation of the replica 1. When
the relative imports that message, its relative replica generation gets set to 1.
When the replica receives an acknowledgment of the data change message, its last
acknowledged generation gets set to 1.
Managing  Distributed  Data   59  
 
 

The Advanced panel displays the following information about generation numbers:

Current Generation — The current generation number, which describes how

many data messages were sent from the replica;

Last Acknowledged Generation — The generation number of the most recent

data message for which an acknowledgment has been received;

Generation (based on last received message) — The current generation number of

the relative replica based on the last message imported into this replica;

The Relative Replica Connection contains the connection information of the relative
replica. You can set this information by browsing to the location of the relative
replica's connection file.
60   Managing  Distributed  Data  
 
 
Replicas created using the workflow for connected environments already know how
to reach the relative geodatabase because the connection files to it was provided at
the replica creation time. In this case for both geodatabases the Relative Replica
Connection in the Replica Manager will already include this information.

Instead replicas created using the workflow for disconnected environment are not
aware of this information. The connection file to reach the relative geodatabase
needs to be provided manually if you want to synchronize data using the workflow
thought for connected environments, otherwise you will only be able to use the
disconnected one.

5.1 Turning a disconnected replica in a connected one

In contexts with limited bandwidth capacity, big latency or internet connectivity not
reliable you may want to create replica using the workflow for disconnected
environments but then be able to synchronize data online, when internet
connectivity allows, and use the synchronization workflow thought for disconnected
environments otherwise.

In order to achieve this you need to provide to your geodatabase information to

reach the relative geodatabase through the Replica Manager.

1. The person who initially created the disconnected replica must add to and
ArcMap document one of the layers in the replica and Open the Replica
Manager tool;

2. Right click the disconnected replica to modify and choose Properties;

Managing  Distributed  Data   61  
 
 
3. Click the Advanced tab, you will notice that the Relative Replica Creation
section is empty. Browse the connection file to the relative geodatabase;

4. Check the Persist user name and password check box. If the user name and
password are not persisted, you will be prompted to provide them every time
you want to run a synchronization online.
62   Managing  Distributed  Data  
 
 
6 Data sharing workflows in WFP

Specific workflows have been designed to facilitate GIS data sharing across WFP
offices; they make use of distributed geodatabases and replica and synchronization
tools to overcome well known networking issues in our organization.

A combination of the different techniques described in the previous chapters is

suggested to allow each office to synchronize data online when Internet connection
is available and reliable, while using disconnected synchronization in case the
online procedure is not feasible because of poor connectivity.

In general we can consider that GIS officers at the WFP headquarters produce and
maintain some datasets that may be of interest for GIS colleagues working in
regional and country offices, such as layers containing information related to
natural hazards (tropical storm tracks, earthquake epicenters, flood extent areas
etc.). At the same time GIS colleagues working in the field collect and maintain
data extremely interested for other GIS practitioners in WFP, such as population
figures, food security and vulnerability data, context related information etc.

All these datasets can be shared across offices setting up one replica for each flow
of information. For instance we’ll create a replica containing all data produced at
the WFP head quarters in which the geodatabase at HQ is the parent and the
destination geodatabase at the CO or RB is the child. One or more replica will then
be created for data sitting and CO or RB level, in this case the roles are inverted,
CO or RB geodatabases are parent and the one at HQ is the child.

Additionally we’ll consider a set of data that is managed at the HQ with a global
extent and a standardized data structure, to which the CO can contribute by editing
existing layers. These include WFP locations of interest (WFP facilities, warehouses)
Managing  Distributed  Data   63  
 
 
and logistics infrastructures (airports, ports, unhas routes, border crossing points
etc.). Such data will be included in an additional replica that from the CO will
populate global layers maintained at the headquarters.

6.1 Data flow HQ à CO and RB

As mentioned before the Geospatial Support Unit at HQ maintains and daily

updates a set of layers related to natural hazards. The one reported below are
considered of interest for country and regional offices, therefore they will be shared
through a specific replica:

o Earthquakes epicenter locations (wld_nhr_eqepicenters_usgs)

o Flood extent areas (wld_nhr_faa30days_dfo)
o Tropical storms buffers, nodes, tracks (wld_nhr_tsbuffers_gdacs,
wld_nhr_tsnodes_jtwc, wld_nhr_tstracks_jtwc)

As these layers are managed at HQ with a global extent, at the replica creation
time spatial filters will be applied according to the spatial extent of the Country
where the destination database is hosted.
This will be a one way replica put in place between the HQ (Parent db) and the
Country Office\Regional Buereau (child DB) using the process for disconnected
environments described in the paragraph 2.2.

After creating the disconnected replica, we need to add the connection file to reach
the destination geodatabase using the steps reported in the paragraph 5.1.
This will enable the replica to be synchronized using the workflow for connected
environments.
64   Managing  Distributed  Data  
 
 
At this step the replica can be synchronized using either the workflow for connected
environments or the one for disconnected environments.
When this replica will be put in place for all Country Offices and Regional Buereaux,
there will be to many replicas to run manual synchronization. For this reason
automatic synchronization workflows will be put in place using the reference
documentation in the paragraph 3.2.5.

6.2 Data flow CO and RB à HQ

In this section we’ll provide some additional information to facilitate the creation of
replicas to send GIS related data from the Country Office\Regional Bureaux to the
WFP Headquarters.

After standardizing GIS datasets and storing them in the Enterprise Geodatabase at
the Country Office, one or more replicas can be put in place to share this data with
other GIS practitioners at the WFP headquarters and regional offices.

The number of replica to create depends mainly to the amount of data available at
the country office and how frequently it is updated.
It is suggested to create group together layers with the same frequency of update,
for instance layers containing administrative boundaries are not updated often
while table containing secondary data maybe be updated on a daily base.
In this case the best strategy would be to create one replica containing all
boundaries datasets, another replica containing all tables, and additional ones
according to update frequency.

All these replicas can be created using a workflow similar to the one described in
the previous paragraph.
Managing  Distributed  Data   65  
 
 
Each of them will be a one way replica where the parent geodatabase is the
Enterprise Geodatabase residing at the country office and the child geodatabase in
this case is the Enterprise Geodatabase at the HQ.
These replicas will be created using the process for disconnected environments
described in the paragraph 2.2.

After creating the disconnected replica, we need to add the connection file to reach
the destination geodatabase using the steps reported in the paragraph 5.1.
This will enable the replica to be synchronized using the workflow for connected
environments.
At this step the replica can be synchronized using either the workflow for connected
environments (automatically or manually) or the one for disconnected
environments.

6.3 Collaborative editing of global layers

Some of the layers used daily at the WFP HQ for mapping or analysis purposes are
managed at a global extent. This set of layers include:

• WFP facilities ( wld_poi_facilities_wfp)

• Warehouses (wld_poi_warehouses_wfp)
• Airports (wld_trs_airports_wfp)
• Ports (wld_trs_ports_wfp)
• Border Crossing Points (wld_poi_bcp_wfp)
• UNHAS routes (wld_trs_unhasroutes_wfp)
• Supply routes (wld_trs_supplyroutes_wfp)
• …..
66   Managing  Distributed  Data  
 
 
The suggested procedure to manage these layers is to adopt a standardized data
structure defined by these global layers at the country office as well.
To achieve these is necessary to put in place a multi-phases process:

1) Share the standardized data structure by creating a 2 ways replica between the
HQ and the CO, which will include all the layers managed at a global extent. At
the replica creation time filters will be applied to replicate only data of interest
for that specific country. Each layer contains in the attribute table an iso3 field;
for instance, when replicating data for Myanmar, definition queries will be
created to replicate features containing only the iso3 = MMR.

2) At the CO, edit replicated layers by filling in information available. For instance
add locations of WFP offices if not already existing or close warehouses that are
not available anymore.

3) Synchronize this data back to HQ using either connected or disconnected

synchronization.