Moving NAV 2009 SP1 from 05 server to 08 r2 server

Are there any known issues with moving NAV from an 05 box to an 08 R2 server, wen running thru citrix.

We have been experiencing multiple problems:

  1. User is logged in and is going about their normal tasks… suddenly an error pops up that the user ID & password is incorrect…NAV crashes

3414.nav.png

Problem 2: Users do not have access to view certain tables that i have verified are in their role, resynced their login … and prior to moving to the R2 server there was no issues. Its like the roles are not syncing all the permissions.

Problem 3: It seems that when a user accesses the tables it is not releasing the table? I have looked in the Sessions tab drilled into the Current Sessions field, the Database Sessions form with Wait Time and Blocking data is blank, and the R2 server has 24 gig of RAM and not experiencing a heavy load at all

5280.nav2.png

Thanks in advance for any insight to any of these problems!

The Dimension question is well known for ages - but it shouldn’t appear only because of SQL version upgrade.
Usual cause - many Analysis views with Update on posting set to YES. You do not need much concurrent and “heavy” docs or thousand-line Journal postings - even a couple of them can block each other.

As for the other part - your post more fits in SQL forum, where our SQL guru Jorg will get it in his sight, so I moved it there. Even my suggestion above about Dims may be wrong…

Thanks for the reply/forum move to get more in depth advice. I understand that table locking can happen if the system is bogged down, if i understand correctly looking at the waits and blocks should show some sign of this, but there has been no blocked transactions or waits that i can see in NAV (or sunning SSMS activity monitor and sql profiler). Our NAV partner did a DB backup of the 05 server and restore on the R2 server, this 08 server has 6 times the resources 24 gig of ram, 30 gig free mem, running on the SAN w/ a fiber channel switch.

–and i forgot to mention table locking happens on more than just this one table… Sales Header, Comment Line, Purchase Header, Dimension-- (tables)

It is still a big mystery why some permissions in a users role are not “taking hold” after the move … Before the move it was working fine, and i have deleted the user and re-added the roles to the user/resync. Other users that have the same role are able to access the tables this particular user does not have perms to access.

As far as NAV crashing and kicking the user out (after they are already in NAV for some time) saying user id and pass are invalid, we believe this might be a Citrix related problem, all our users access NAV thru Citrix.

That’s about where i am today on these issues

Hi!

Well, there are basically no issues with upgrading from SQL 2005 to 2008 R2! Of course, the SQL 2008 engine behaves differently, but actually in an improved way - still there is a smal chance that certain queries perform slower than before … means: as any SQL Server also that one requires some tuning …

As for the Logins: the Database site Users were transferred via the DB Backup, but the SQL Instance Logins have to be transferred manually:
http://support.microsoft.com/kb/246133

Blocks and Deadlock occur in any NAV/SQL environment; if you experience a different behavior you’ll need to investigate this in detail:
http://dynamicsuser.net/blogs/stryk/archive/tags/Locking/default.aspx

One important thing which is sometimes forgotten when upgrading the SQL Server: you have to change the “Database Compatibility Level” to 100 (SQL 2008) - else your database only “understands” SQL 2005 syntax, thus it cannot take advantage of the full SQL 2008 stack!

Thanks for the reply, I will investigate these links.

We nailed down one problem to citrix, NAV crashing…

I ran the optimizer in the File–> database → information… and scheduled a regular SQL job to reindex the NAV server…

Some Users are still experiencing hiccups with the permissions in their roles not taking hold…

The “Table Optimizer” actually recreates all indexes (CREATE INDEX … WITH DROP_EXISTING) while the SQL Task “Index Rebuild” runs an ALTER INDEX … REBUILD command. In both cases the idea is to defragment indexes, but as TabOpt recreates all from the scratch it is causing super load on the system, AlterIdx is somewhat “smoother” … I’d like to refer to my BLOG about more details about indexing:
http://dynamicsuser.net/blogs/stryk/archive/tags/Indexes/default.aspx

According to this it is crucial to run periodic Statistic Updates, e.g. via MP job. Once you take care about stats I have quite good experiences with these DB settings:

Auto Create Stats = FALSE
Auto Update Stats = TRUE
Auto Update Async = TRUE