EV Client Monitoring

[Simon_Butler]

A recent issue with cache downloads has highlighted the fact that we have no monitoring to show whether laptops are successfully downloading the vault cache.  I guess the same rings true for meta cache for all clients.

Has anyone come up with a solution that will allow pro-active monitoring of clients (rather than waiting for end-users to find missing items and/or open the sync properies and see an error and raise helpdesk tickets...)?

If not - are there any hooks on the client that we could use to develop our own monitoring?

fyi - no server event log type errors relating to our current download issue, so cannot monitor from that perspective.

Cheers,

Si

[GertjanA]

Hi Simon,

On the Vault-server, log in as VaultService Account.

open Internet Explorer, then goto

localhost/enterprisevault/vcview.aspx

You see there an overview of VC's being build, being downloaded, being completed.

If that is usefull, I do not know. I ran into this by accident.

[Rob_Wilcox1]

You can also construct some queries against the IIS logs.  Check the articles section, in the forums, which should give you some ideas.

If there is anything specific you'd like the article to include, let me know..

[M__Urban]

I completely agree with this post.  We have found that in many cases the synchronization just stopped for an end-user with no notification.  Then we find hundreds, and in some cases thousands of emails the user has moved to their virtual vault that never made their way up to the server.  The vcview and iis is no solution whatsoever for this type of problem.

[JesusWept3]

Then simply either stop users from moving items around virtual vault and in to virtual vault until such a time as you feel the product is more robust and meets your needs, or talk to your sales representatives and open an idea in the Ideas section

Or perform some kind of user education, after all its the same with outlook in many respects, you don't get any kind of alert when an end users fails to build an OST or their PST corrupts etc etc, its something they find out for themselves

[Rob_Wilcox1]

In part I agree with JW2, and in part I agree with M. Urban.  What I'd like to see/discuss/elaborate on is :

What do we want to do/see ... in an ideal world?

My thoughts are from two different angles ...  there is the possibility that a user hasn't checked-in with the server for a long period of time, and has a large build of messages on the client - moves and so on.  For these we need someway of notifying the server, so it can be displayed in a console view, report, email, or whatever.  The other possibility is that the server has a lot of data hanging around waiting fo the client to hook up.   I think key to some of this is changes that would be need to made to the mailbox synchronisation process -- nuts and bolts trying to discuss "solutions", I know.. but I'm trying to figure out REALISTICALLY how this can be done, and what sort of problems people see that need to be flagged to either an end-user, or an administrator or both.

I mean, for example, if a user opens Outlook or has Outlook open after a long period of time (a week, 2?) and the modified date on their hidden message, containing their settings, is "old", that should be something that the end-user is told....  and an admin is told (somehow).

[JosephRodgers]

All,

We have had great success using the VVAutoSyncItemThreshold & VVAutoSyncSizeThresholdMB reg keys (these are desktop policies settings with EV 9.  We are running EV 8 servers with 9.0sp2 client).  With these keys set to reasonable levels (we use 10 \ 5MB) it will make the client vaults start the sync much sooner and (hopefully) avoid any major build up on client side.

Ideal world would be a client side checkin with relevent info.  No client checkin might require investigation (SEP does something similiar).

Simplier solution might be to write to client event log which can then be polled via centralized management utility.

-Joe

[Simon_Butler]

Hi - Thanks for the replies.

VCView - Have used that on an ad-hoc basis & useful for support queries.  I'm going to reverse engineer to create something that will span multiple servers & multiple directories/sites.

IIS Logs - I'll take a look at the posts.  Sounds like I'd need a list of success type requests and errors - perhaps based on 40x and other responses.

"Client Check-in" is the sort of thing I'm after.  I think at a minimum we need a report which shows the last successful initial or incremental sync time for each enabled mailbox/archive.  If possible, client reported errors should be posted back to the server, i.e. the ones seen in the sync window in Outlook.  I realise in some cases that may not be possible as connectivity may be the cause, but for scenarios where corrupted MDC/DBs cause sync failures this should be possible.

Cheap & nasty would be an option for us to enable better feedback to end-user.  If there is a sync issue, prompt the user....configurable with friendly message to call the helpdesk...  I hate that, but if that was an easy temp fix I'd take it.  Having the issue hidden and only found when the user is offline using the cache is not a great service...

Cheers,

Si