Alfresco lucene Index

[itayh]

Hi,

I have several questions about lucene in alfresco.
I could not find documentations on those issues.

1 - What is archive folder under \lucene-indexes means?
2 - Can I set <index enabled="false"> for specific items and then they will not be index
3 - What <archive>true</archive> in cm:content mean?
4 - basically I want to reduce the size and number of files/folders under lucene-indexes.

Any Ideas?

Thanks a lot in advance,
Itay

[itayh]

Also what backup-lucene-indexes, indexes-backup means?

I will love to get a link to the documentation of those indexes.

[itayh]

Hi,

We also set the indexBackupJobDetail to run in the middle of the day. We notice something very strange: As long the backup was running we could NOT connect to alfresco. The backup took ~ 90 minutes!

Is it make any sense?

Its look like we will need to disable this job. What will be the impacts of this?

Thanks in advance,
Itay

[andy]

Hi

The back up job should not take anything like that long. It just copies the indexes - it does hold a lock on individual indexes so they can not change. This will be much improved soon.

It is up to you if you need a back up - it can make rebuilding an index faster - you can always rebuild from nothing.

1) There are indexes for all store types and stores - archive just happens to be the store type where archived nodes goes.

2) Changing DDs settings only changes indexing behaviour for each attribute - some stuff will always be indexed.

3) Archiving is what may happen to nodes when you delete them - set to false if you do not want this. They are recoverable from the archive and will be indexed.

4) Backup indexes are created by the back up job - you can delete old ones you do not want - or have non at all

Andy

[itayh]

Hi,

Thank you for your answer.

Can you please tell me what are the differences between recovery with backup index and not (I mean the estimated time difference. How fast is backup with index compare to without).

What is the recommended way to do (with backup or without)?

Also if I set <archive>false</archive> in cm:content and cm:folder will it be ok with system functionality?

Itay

[itayh]

Hi,

I notice that just copy the indexes takes a lot of time (we have a lot of indexes).

Also If I run the SDK FirstFoundationClient program that uses Alfresco Embedded project to create nodes, Will it create indexes? I add many nodes using FirstFoundationClient  and could not see any indexes.

Thanks a lot in advance,
Itay

[andy]

Hiu

Can you post your index structure (file names, directories, structure and sizes) and repo version OS etc.

Where are you indexes stored?

The SDK may be storeing data where you do not expect depening on the repository properties settings - in the root folder of the project being run from eclipse.

Andy

[itayh]

Hi,

Those are the indexes that were created using java SDK program (we add many nodes):
# du -sch /opt/alfresco/host-specific/node-7/indexes{,-backup}/*
3.8M    /opt/alfresco/host-specific/node-7/indexes/archive
4.0K    /opt/alfresco/host-specific/node-7/indexes/locks
172K    /opt/alfresco/host-specific/node-7/indexes/system
40K     /opt/alfresco/host-specific/node-7/indexes/user
362M    /opt/alfresco/host-specific/node-7/indexes/workspace
1.7M    /opt/alfresco/host-specific/node-7/indexes-backup/archive
4.0K    /opt/alfresco/host-specific/node-7/indexes-backup/locks
132K    /opt/alfresco/host-specific/node-7/indexes-backup/system
40K     /opt/alfresco/host-specific/node-7/indexes-backup/user
83M     /opt/alfresco/host-specific/node-7/indexes-backup/workspace
450M    total

Those are the indexes that were created using alfresco (we add less node than in SDK):
# du -sch /opt/alfresco/host-specific/mg1/indexes{,-backup}/*
56M     /opt/alfresco/host-specific/mg1/indexes/archive
4.0K    /opt/alfresco/host-specific/mg1/indexes/locks
632K    /opt/alfresco/host-specific/mg1/indexes/system
224K    /opt/alfresco/host-specific/mg1/indexes/user
555M    /opt/alfresco/host-specific/mg1/indexes/workspace
56M     /opt/alfresco/host-specific/mg1/indexes-backup/archive
4.0K    /opt/alfresco/host-specific/mg1/indexes-backup/locks
632K    /opt/alfresco/host-specific/mg1/indexes-backup/system
224K    /opt/alfresco/host-specific/mg1/indexes-backup/user
555M    /opt/alfresco/host-specific/mg1/indexes-backup/workspace
1.2G    total


We use linux as OP
Linux mg1 2.6.18-4-amd64 #1 SMP Fri May 4 00:37:33 UTC 2007 x86_64 GNU/Linux

Also Can you please tell me what are the differences between recovery with backup index and not (I mean the estimated time difference. How fast is backup with index compare to without).

What is the recommended way to do (with backup or without)?

Also if I set <archive>false</archive> in cm:content and cm:folder will it be ok with system functionality?

Thx a lot in advance,
  Itay

[andy]

Hi

Can you post the listing of directories in workspace/Spaces?
(The index info would also help)

The index rebuild difference depends on which version you are running. If it does a full rebuild, having the index can be slower and you may be better off deleting it, if it is a partial rebuild then it can be much faster (as it validates stuff is present). 

Not archiving is up to you - there is no support to undelete via the UI. When you delete it has gone.

Andy

[itayh]

Hi,

The Index backup job taking more than an hour. This is the most important issue that we try to understand.During this time everything is block. We remove this job for now but we want to understand why it happen.

The index.recovery.mode=VALIDATE in repository properties.

How can we investigate it?

Thx,
  Itay