Custom Filtering by words, or phrases in message body. Retention Folders and SQL Replication.

[wandarah]

Me Again, 

Stupid question guy!

I've got a client at the moment that believe they have a requirement for a couple of things (I don't have access to test these at the moment, painfully enough) that aren't explicitly mentioned in the documentation, nor does anyone I usually consult with know the answer. I'd again, appreciate any advice from the community. 

1. They want to create retention folders in users Outlook. That's fine, we can do that using EVPM - and indeed, create as many folders as we like, or script a set of folders for a particular job role, or 'persona'. All of that is fine - what I cannot see mentioned for Exchange Mailboxes however, is this - do subfolders of a retention folder created by EVPM and added by the user inherit the permissions of the parent folder? I know this is the case, or can be, for Public Folders. Using Exchange 2010 Managed Folders is out of the question unfortunately. 

2. In addition to user-driven classification of emails using the above retention folder system mentioned - they want a secondary action of an Archiving Task to be the use of custom filters - i.e scan eligible items for archiving and match them against rule sets, which would either move them to the rentention folders mentioned, or not depending on the action assigned to the rule set - my question is, can custom filters be used to scan the 'message body' for particular keywords or phrases? I know you can do so on subject, etc. Or is this a job that could only be managed by the ACE, etc. Even if this can be done, I assume I would need at least two archive tasks pointing to the same target....yes?

3. A quick one - I assume this is still the case, but can EV9 still not talk to both sides of an SQL replication partnership? i.e In a centralized deployment, over two Datacentres, running SQL Clusters in each - the EV servers in either centre cannot talk to the local SQL cluster and have the replication between them handle the management of the DB's. They would need to use one cluster in one of the datacentres for the EV Site, and use the other cluster as a form of DR. 

Any advice much appreciated. 

[JesusWept3]

So to answer the first question

If i enable user A and give a blanket policy and retention for the entire archive, then i use EVPM to create \My 7 Year Folder\ and slap a 7 year retention on it, anything underneath will inherit the same things

So for instance i could do \My 7 Year Folder\SubFolder\SubFolder\RE: My Message
That message would have the 7 year subfolder

Basically EV Will trawl its folder, and the parents folder and the parents parents folder until it finds a "Filter" Message that changes the defaults of what is set on the mailbox

If it does not find any such Filter message, then it will use the default settings on the mailbox. The Filter messages are set either by the user changing the settings themselves on the Enterprise Vault Tab (you would have had to unlock your policies to do this and they would have to have the DCOM client) or by EVPM.

And note that if you ever ZAP the mailbox to remove properties of EV from the mailbox, it will remove those filter messages, so you may suddenly find that users are putting items in to \my 7 year folder\ but getting a 1 year retention category.

So make sure you note as to what EVPM scripts are running against who, that way if you do need to zap the mailbox for any number of reasons, you can be sure to apply the same EVPM and filter settings that once had

For the second question I don't believe Enterprise Vault out the box has that kind of functionality to move messages about dependant on their content, and even with custom filters i believe the only place you can move items to is the deleted items folder and what not.

Rob Wilcox wrote a fairly good article on beginning with Custom Filters and there is more obviously in the admin guides too

http://aka-community.symantec.com/connect/articles/custom-filtering-enterprise-vault

 

And question #3, in a cluster you would just target the virtual name right? and when it fails over the virtual name points to the second physical node as opposed to the first so there really shouldn't be an issue there

EV itself will not try and discover what the active SQL Server is, it just as several DSN's and connection strings that just point to servernames and isn't aware when hardware failovers take place

[wandarah]

Thanks, 

1. Yep - will test when I can obviously, but that's what I expected. This means that users can create folders till their hearts content and it'll take the retention category of whatever the 'root' EVPM created folder is. Good. 

2. Damn, really? I had assumed the Action 'Move_Deleted_Items' rule could be modified to any folder name. Have read guide by Rob, not quite the same sorta thing we're going for here though. 

I need to know if we can apply retention, etc based on keywords in the message body - not just the subject, etc. 

3. Yeah, that's true. I'm more talking about two seperate clusters with bi-directional replication. So, EVServer1 writes to SQLCLUSTER1, and EVServer2 writes to SQLCLUSTER2 - and the two clusters maintain a single site. 

[JesusWept3 · Accepted solution]

#2 : Am absolutely sure, however i think there may be some feature enhancements forthcoming soon that may help you out better, but i don't have much detail on it

and #3: OK I get what you're saying, i guess you could change the EnterpriseVaultDirectoryDSN and other DSN's to point at different SQL Servers on different EV Servers, would not recommend it though as that would be a cheap hack

The only real thing you can do with seperate SQL Servers is have a vault store on different SQL Servers and what not but thats not what you're after either.

[wandarah]

Farrrrrrr, thanks man.