none
Re-index Exchange database copy

    Question

  • I have installed the PDF iFilter for Exchange 2010, and have registered it.

    I now need to re-index the databases using the ResetSearchIndex.ps1 script, however, when I do that against the live DB, it makes the database unsearchable during the re-index. 

    The last database I re-indexed took ~23 hours for a 700GB database. 

    I have 34 Databases total spread across 6 Exchange servers.  Largest database I have is 1TB.

    I am considering running the re-index on the secondary databases in my DAG, then when they are complete I can activate the secondary database.

    After that, I can re-seed the primary DB index from the secondary database and re-activate it when finished.

    Does that sound reasonable? Or will I have problems activating the secondary after it's been crawled since it doesn't match the primary's index?

    This might explain it better

    Action Sequence

    Copy1

    Copy2

    LAG

    1

    Active - Users Accessing

    Re-Index (Crawling)

    No Action

    2

    Re-Seed Index From 2

    Active –Using New Index

    Re-Seed Index from 2

    3

    Re-Activate – Using New Index

    Passive Again

    No Action


    Bobby Pendino


    • Edited by BPendino Sunday, June 16, 2013 4:51 AM
    Sunday, June 16, 2013 4:40 AM

Answers

  • This not going to give you much Bobby.  The passive copy is indexing the active copy over the MAPI network and then saves the CI files locally.

    This article was updated previously to better explain the process

    http://technet.microsoft.com/en-us/library/bb232132(v=exchg.141).aspx#DAG

    In organizations that have a database availability group (DAG), during the seeding process, DAG members with a passive mailbox database copy replicate the content index catalog from the DAG member that has the active mailbox database copy. The content index is typically 10 percent the size of the mailbox database. After initial seeding, the server with the passive database copy gets message data from the server with the active database and performs content indexing locally. The bandwidth used for copying message content for indexing is in addition to the bandwidth used for replication of transaction logs. When planning a high availability deployment, you must consider the bandwidth used by Exchange Search.

    When you are indexing, what disk response times do you get? 

    Also what Exchange build are you on please?


    Cheers,

    Rhoderick

    Microsoft Senior Exchange PFE

    Blog: http://blogs.technet.com/rmilne  Twitter:   LinkedIn:   Facebook:   XING:

    Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

    • Marked as answer by Zi Feng Saturday, June 22, 2013 8:48 AM
    Sunday, June 16, 2013 1:17 PM
  • So, you are correct in this fact: "The passive copy is indexing the active copy over the MAPI network and then saves the CI files locally."

    However, that did not prevent me from creating a new index on the passive copy, so it did help me since my goal was to keep the live database searchable for clients while I re-indexed the passive copy.

    Here's what I ended up with:

    Copy1 (Active) Index State

    Copy2 (Passive) Index State

    Copy3 (LAG) Index State

    Healthy/Searchable

    Crawling

    Healthy

    Tonight I plan to Activate the Secondary copy and then re-seed the Primary and LAG Copy indexes from Copy2.  This will keep the Databases searchable for all of my clients and re-index the databases to capture pdf information.

    Tonight my scenerio will be:

    Copy1 (Passive) Index State

    Copy2 (Active) Index State

    Copy3 (LAG) Index State

    Re-Seeding

    Healthy

    Re-Seeding


    Bobby Pendino

    • Marked as answer by BPendino Friday, June 28, 2013 2:42 PM
    Friday, June 28, 2013 2:42 PM

All replies

  • This not going to give you much Bobby.  The passive copy is indexing the active copy over the MAPI network and then saves the CI files locally.

    This article was updated previously to better explain the process

    http://technet.microsoft.com/en-us/library/bb232132(v=exchg.141).aspx#DAG

    In organizations that have a database availability group (DAG), during the seeding process, DAG members with a passive mailbox database copy replicate the content index catalog from the DAG member that has the active mailbox database copy. The content index is typically 10 percent the size of the mailbox database. After initial seeding, the server with the passive database copy gets message data from the server with the active database and performs content indexing locally. The bandwidth used for copying message content for indexing is in addition to the bandwidth used for replication of transaction logs. When planning a high availability deployment, you must consider the bandwidth used by Exchange Search.

    When you are indexing, what disk response times do you get? 

    Also what Exchange build are you on please?


    Cheers,

    Rhoderick

    Microsoft Senior Exchange PFE

    Blog: http://blogs.technet.com/rmilne  Twitter:   LinkedIn:   Facebook:   XING:

    Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

    • Marked as answer by Zi Feng Saturday, June 22, 2013 8:48 AM
    Sunday, June 16, 2013 1:17 PM
  • So, you are correct in this fact: "The passive copy is indexing the active copy over the MAPI network and then saves the CI files locally."

    However, that did not prevent me from creating a new index on the passive copy, so it did help me since my goal was to keep the live database searchable for clients while I re-indexed the passive copy.

    Here's what I ended up with:

    Copy1 (Active) Index State

    Copy2 (Passive) Index State

    Copy3 (LAG) Index State

    Healthy/Searchable

    Crawling

    Healthy

    Tonight I plan to Activate the Secondary copy and then re-seed the Primary and LAG Copy indexes from Copy2.  This will keep the Databases searchable for all of my clients and re-index the databases to capture pdf information.

    Tonight my scenerio will be:

    Copy1 (Passive) Index State

    Copy2 (Active) Index State

    Copy3 (LAG) Index State

    Re-Seeding

    Healthy

    Re-Seeding


    Bobby Pendino

    • Marked as answer by BPendino Friday, June 28, 2013 2:42 PM
    Friday, June 28, 2013 2:42 PM