1. The GP 10 year end update/service pack 3 upgrade process through utilities can be run from any machine, not just the server (although it may be quicker on the server).
2. Hedge your bets, make a backup of the reports.dic before applying GP 10 SP3/year end update and then rebuild the reports dictionary after applying.
3. Remember, your user date in Dynamics GP has to be in the new year to do a check run after closing 2008.
4. Backup, backup, and oh, yeah, backup. And keep them...one from pre-year end, and one from post-year end.
5. Allow enough time to apply the updates, AND check your year end wage file.
6. In a crunch, you can let one person in to run a payroll without applying the updates to other workstations. Of course, the un-updated workstations cannot be used until the update is applied. But this will help you in a crazy pinch. And, I know, I know, it is contrary to what the guide tells you.
Now, if I can only manage to remember these little tidbits NEXT YEAR :)
Tuesday, January 6, 2009
"Unable to load SanScript Library" and Integration Manager
Late last month, I was asked to assist a client who was having issues with the year end update for Dynamics GP. After I had helped her through the issue, she mentioned that she had an issue with Integration Manager that her previous consultant could not resolve. She had even gone so far as to completely wipe and reload the machine in an attempt to address the issue. Naturally, in the middle of a million year end questions, this issue actually sparked my curiousity :)
She showed me that on one machine, if she tried to run an integration, it would appear to start and then would display the error "Unable to load SanScript Library". Like any good consultant, I immediately checked the KnowledgeBase and found a few articles that discussed the following possible causes and resolutions...
1. Corrupt install or corrupt IM.dic: Uninstall, clean registry, and then reinstall
2. Corrupt integration: Recreate integration
I knew she had been through #1 several times, so I tried #2 and still received the error. I tried using a backup IM.MDB, and still received the error. Hmmm.
So I started to think about the corrupt IM.dic mentioned in the article. And I started to think about how many of the articles mentioned the error when opening an integration, but she did not the error until she actually ran the integration (and that it appeared to run at first before displaying the error). It was as if it was having issues with pushing the data through the Dynamics GP client, as opposed to an issue with the Integration Manager application itself. And then I remembered that she launched Integration Manager from the start menu, because the option from Tools>>Integrate was not available to her.
I bet many of you already know where I am heading with this. I checked her Dynamics.set file (her launch file, through Tools>>Setup>>System>>Edit Launch File) and, sure enough, Integration Manager was not included in the launch file. The IM.dic was were it should be, in the GP directory, but the Dynamics.set was not even looking for it. We modified the launch file to include Integration Manager (by copying from the Dynamics.set on a working workstation), and the integration now runs perfectly.
This was on GP 9, not sure if others have run in to it on GP 10. But I thought it was an interesting find that I should share with others, to spare you the troubleshooting should you come across it. Happy New Year!
She showed me that on one machine, if she tried to run an integration, it would appear to start and then would display the error "Unable to load SanScript Library". Like any good consultant, I immediately checked the KnowledgeBase and found a few articles that discussed the following possible causes and resolutions...
1. Corrupt install or corrupt IM.dic: Uninstall, clean registry, and then reinstall
2. Corrupt integration: Recreate integration
I knew she had been through #1 several times, so I tried #2 and still received the error. I tried using a backup IM.MDB, and still received the error. Hmmm.
So I started to think about the corrupt IM.dic mentioned in the article. And I started to think about how many of the articles mentioned the error when opening an integration, but she did not the error until she actually ran the integration (and that it appeared to run at first before displaying the error). It was as if it was having issues with pushing the data through the Dynamics GP client, as opposed to an issue with the Integration Manager application itself. And then I remembered that she launched Integration Manager from the start menu, because the option from Tools>>Integrate was not available to her.
I bet many of you already know where I am heading with this. I checked her Dynamics.set file (her launch file, through Tools>>Setup>>System>>Edit Launch File) and, sure enough, Integration Manager was not included in the launch file. The IM.dic was were it should be, in the GP directory, but the Dynamics.set was not even looking for it. We modified the launch file to include Integration Manager (by copying from the Dynamics.set on a working workstation), and the integration now runs perfectly.
This was on GP 9, not sure if others have run in to it on GP 10. But I thought it was an interesting find that I should share with others, to spare you the troubleshooting should you come across it. Happy New Year!
Wednesday, December 24, 2008
Inline Scalar Function Performance (T-SQL geeks only)
If you are technical and work alot with SQL Server, I highly recommend subscribing to SQL Server Magazine. One of the contributing authors, Itzik Ben-Gan, is a T-SQL expert in the extreme.
Itzik recently posted an excellent article about performance issues using inline scalar functions (functions that return only 1 value and are used within a larger query). He demonstrates the performance of a typical inline scalar function, and then offers a relatively simple alternative approach that dramatically improves performance.
http://www.sqlmag.com/articles/index.cfm?articleid=101104&
In his sample, he demostrates how this change makes his 14 second query run in less than 1 second. That's a massive performance improvement for changing just a few keywords that will not affect the structure of your query.
Since I wrote an earlier post about UDFs, I thought that this would be a great follow up. To summarize his post:
1) Add the "RETURNS TABLE" clause to your UDF
2) Change your inline function call to use a SELECT statement against the function
Well worth the minimal effort, even if you aren't querying a million rows.
Itzik recently posted an excellent article about performance issues using inline scalar functions (functions that return only 1 value and are used within a larger query). He demonstrates the performance of a typical inline scalar function, and then offers a relatively simple alternative approach that dramatically improves performance.
http://www.sqlmag.com/articles/index.cfm?articleid=101104&
In his sample, he demostrates how this change makes his 14 second query run in less than 1 second. That's a massive performance improvement for changing just a few keywords that will not affect the structure of your query.
Since I wrote an earlier post about UDFs, I thought that this would be a great follow up. To summarize his post:
1) Add the "RETURNS TABLE" clause to your UDF
2) Change your inline function call to use a SELECT statement against the function
Well worth the minimal effort, even if you aren't querying a million rows.
Thursday, December 18, 2008
Dynamics GP and SharePoint Resources
SharePoint is a popular discussion item with clients lately; it seems to come up alongside discussions of Business Portal and deploying Dynamics GP information to a larger group of users (who do not necessarily have access to Dynamics GP directly). So I thought I would share some of the resources I use regularly as support in these discussions. Please share your popular links as well, and I will update the blog :)
SharePoint Edition Comparison, to understand differences between Windows SharePoint Services (WSS) and Microsoft Office SharePoint Server (MOSS), along with other editions:
http://office.microsoft.com/en-us/sharepointtechnology/FX101758691033.aspx
For Dynamics Partners, information on the Dynamics Client for Office. The DCO is a bundled licensing of Microsoft Office SharePoint Server and other Dynamics features. Includes a FAQ, Fact Sheet, and Presentation:
https://mbs.microsoft.com/partnersource/newsevents/news/newsgeneral/microsoftdynamicsclientofficesharepoint.htm?printpage=false
High level information on the Dynamics GP and Microsoft Office SharePoint Server integration points:
http://www.microsoft.com/dynamics/gp/product/sharepointofficeintegration.mspx
For Dynamics Partners, fact sheets on specific integration points between MOSS and Dynamics GP:
SharePoint Edition Comparison, to understand differences between Windows SharePoint Services (WSS) and Microsoft Office SharePoint Server (MOSS), along with other editions:
http://office.microsoft.com/en-us/sharepointtechnology/FX101758691033.aspx
For Dynamics Partners, information on the Dynamics Client for Office. The DCO is a bundled licensing of Microsoft Office SharePoint Server and other Dynamics features. Includes a FAQ, Fact Sheet, and Presentation:
https://mbs.microsoft.com/partnersource/newsevents/news/newsgeneral/microsoftdynamicsclientofficesharepoint.htm?printpage=false
High level information on the Dynamics GP and Microsoft Office SharePoint Server integration points:
http://www.microsoft.com/dynamics/gp/product/sharepointofficeintegration.mspx
For Dynamics Partners, fact sheets on specific integration points between MOSS and Dynamics GP:
- General Fact Sheet on MOSS and Dynamics GP integration points, https://mbs.microsoft.com/partnersource/sales/salestools/productfactsheets/MD_GP10_SP_Server
- For Search, https://mbs.microsoft.com/partnersource/sales/salestools/productfactsheets/SearchGP10
- For Workflow, https://mbs.microsoft.com/partnersource/sales/salestools/productfactsheets/MSDYGP10_WorkflowFS
- For Business Portal, https://mbs.microsoft.com/partnersource/sales/salestools/productfactsheets/MSDYGP10_BusnPortalFS.htm
Project Accounting Default Account Sourcing
Currently I am working on several project accounting implementations. When I have several similar projects going on at the same time like this, I tend to focus on the common themes. So, lately, I have been thinking about the default project account setup through Tools>>Setup>>Project>>Project>>Accounts. This is always an interesting thing to explain to clients when they ask, how does project determine what accounts to use? And my answer is, normally, it depends on your setup.
In the core modules of Dynamics GP, the defaulting of posting accounts is fairly straightforward. The system looks first to a master record, and if no accounts are specified, it then looks to the posting account setup. So, for example, when entering a payables transaction, the system will first look to the vendor card for default posting accounts and whatever it cannot find there, it will attempt to pull from the posting accounts setup (Tools>>Setup>>Posting>>Posting Accounts).
Project accounting presents users with a more flexible defaulting model, where default locations can be specified by transaction type and by account. Additionally, segment overrides can also be specified at the contract and project levels. This creates tremendous flexibility in terms of how the system can default accounts, and where accounts need to be set up in the first place.
So, to outline the different options...
For each cost transaction type in project accounting:
In the core modules of Dynamics GP, the defaulting of posting accounts is fairly straightforward. The system looks first to a master record, and if no accounts are specified, it then looks to the posting account setup. So, for example, when entering a payables transaction, the system will first look to the vendor card for default posting accounts and whatever it cannot find there, it will attempt to pull from the posting accounts setup (Tools>>Setup>>Posting>>Posting Accounts).
Project accounting presents users with a more flexible defaulting model, where default locations can be specified by transaction type and by account. Additionally, segment overrides can also be specified at the contract and project levels. This creates tremendous flexibility in terms of how the system can default accounts, and where accounts need to be set up in the first place.
So, to outline the different options...
For each cost transaction type in project accounting:
- Miscellanous Logs
- Equipment Logs
- Purchases/Materials
- Employee Expenses
- Timesheets
And then for each account type (I only list a few of the basics below for the example that follows):
- Cost of Goods Sold/Expense
- Contra Account for Cost
- Project Revenue/Sales
- Accounts Receivable
For each account type, you can specify a "source" to pull the default accounts from:
- None (do not create distributions for this account)
- Customer (use Cards>>Sales>>Customer>>Project>>Accounts)
- Contract (use Cards>>Project>>Contract>>Accounts)
- Project (use Cards>>Project>>Project>>Accounts)
- Cost Category (use Cards>>Project>>Cost Category>>Accounts)
- Trx Owner (depending on transaction type, can pull from Employee, Equipment, Miscellaneous, or Vendor)
- Specific (enter one specific account to always be used)
- The expense account used on timesheets should vary by the cost category, but the project determines which department segment should be used.
- The offset to the expense (the contra account for cost) is determined by the employee's home department.
- When the labor is billed, the revenue account should vary by the cost category, but the project determines which department segment should be used.
- When the labor is billed, the AR account will be different based on the customer.
Sometimes it takes a bit of discussion to get items distilled down to the bullet points I show above. I actually start with a cost category worksheet that I have clients complete as homework that documents the cost categories and accounts they need in project accounting as a way to help us think through the requirements. Then, from that worksheet, we can discuss how to best set up the account sourcing. Feel free to post on this blog and I am happy to share the worksheet I use in case anyone is interested.
So, the outcome we came to for timesheets in the example above:
- Cost of Goods Sold/Expense- Source: Cost Category
- Contra Account for Cost- Source: Trx Owner
- Project Revenue/Sales- Source: Cost Category
- Accounts Receivable- Source: Customer
Then, at the project level, we will specify a sub-account format (Cards>>Project>>Project>>Sub Account Format Expansion Arrow>>Account Segment Overrides) that includes the department segment value and we will mark to override the segment on the Cost of Goods Sold/Expense and Project Revenue/Sales account. So this setting will substitute in the department value specified on the project, while pulling the remaining account segments from the cost category.
With the sources set as noted above, the client needs only specify accounts as follows:
- On the cost category (Cards>>Project>>Cost Category>>Accounts), enter the Cost of Goods Sold/Expense and Project Revenue/Sales accounts.
- On the employee (Cards>>Payroll>>Employee>>Project>>Accounts), enter a Contra Account for Costs.
- On the customer (Cards>>Sales>>Customer>>Project>>Accounts), enter an Accounts Receivable account for timesheets.
And, more importantly, the accounts do not have to set up repeatedly in multiple places. And, because the project determines the department segment, there is no need for cost categories to be set up for each department, only for each main account needed.
Although the effort involved in coming to this conclusion may make you feel like accepting the default of all accounts defaulting from the cost category, taking the time to set this up in the most efficient way possible will minimize mistakes in posting and simplify the setup in the long run.
As always, please feel free to share your experiences and perspectives. Happy holidays!
Monday, December 15, 2008
More HyperV Resources
Today I tried in earnest to migrate a Virtual Server 2005 VHD file to HyperV. Things looked good until I realized the networking in the virtual server did not work. That's when I found this article that offers some steps on preparing a virtual server before you move the VHD to HyperV. Of course, I would recommend making a backup of your VHD and VMC before making these changes, just in case.
http://blogs.technet.com/askcore/archive/2008/09/19/migrating-to-hyper-v-from-virtual-server-or-virtual-pc-tips-and-suggestions.aspx
Which led to this tool, which is apparently a simple means for translating the VMC file to HyperV. Since I only have a few servers to migrate, I don't know that this tool is particularly valuable to me--it only takes a few minutes for me to create a new virtual server and then just attach the VHD. If you have alot of servers to migrate, the article mentions that System Center Virtual Machine Manager is probably a better option.
http://blogs.technet.com/matthts/archive/2008/09/12/vmc-to-hyper-v-import-tool-available.aspx
And finally, here is a post with an amazing list of HyperV links and resources. I will be looking over these during the next few weeks.
http://blogs.technet.com/virtualworld/archive/2008/11/24/pfe-hyper-v-ops-day-october.aspx
http://blogs.technet.com/askcore/archive/2008/09/19/migrating-to-hyper-v-from-virtual-server-or-virtual-pc-tips-and-suggestions.aspx
Which led to this tool, which is apparently a simple means for translating the VMC file to HyperV. Since I only have a few servers to migrate, I don't know that this tool is particularly valuable to me--it only takes a few minutes for me to create a new virtual server and then just attach the VHD. If you have alot of servers to migrate, the article mentions that System Center Virtual Machine Manager is probably a better option.
http://blogs.technet.com/matthts/archive/2008/09/12/vmc-to-hyper-v-import-tool-available.aspx
And finally, here is a post with an amazing list of HyperV links and resources. I will be looking over these during the next few weeks.
http://blogs.technet.com/virtualworld/archive/2008/11/24/pfe-hyper-v-ops-day-october.aspx
Sunday, December 14, 2008
Transitioning to HyperV
For the last few years, I have used Virtual Server 2005 to host virtual machines for testing different Dynamics GP environments, and for hosting different development environments. Some clients still use SQL 2000 and GP 8, while others are using SQL 2008 with the latest GP 10 service pack. Maintaining those configurations on a single physical machine typically isn't practical, so Virtual Server offered a great way to easily maintain and backup multiple server configurations. Virtual Server uses the same VHD files as Virtual PC, but I found that it was more convenient to use, easier to manage VMs, and seemed to be faster than Virtual PC.
I've finally built a new Windows 2008 Server, so I'm going to be transitioning my virtual machines over to HyperV, which is the replacement for Virtual Server 2005. Because of the memory requirements of virtual machines, I've installed the 64-bit (x64) version of Server 2008, which allows me to get past the 3GB / 4GB limit of 32-bit windows. The desktop machine I used for Server 2008 only supports up to 8GB of RAM, but I figure that should allow me to pretty comfortably host up to 4 virtual servers (typically using 1.5GB of memory each)--more than I need running at any one time.
I'm still learning about HyperV and how it is different from Virtual Server 2005, and in which ways it is the same, so for the time being I'll offer some links to articles that found helpful.
My focus is not just learning and understanding the HyperV features, but also understanding how to perform equivalent maintenance operations and processes in HyperV to replace the tricks I learned with Virtual Server 2005. For instance, what is the best way to backup a HyperV image? The files that are created by HyperV are slightly different than VS 2005, so I need to learn what to backup, when, and how (i.e. Can I write a script to pause the VM and run a backup like I did with VS 2005?). Also, can I migrate an existing VS 2005 VHD file to HyperV (the answer is yes), and if so, what steps do I need to complete to get all of the server features working properly in HyperV? (still working on this one...)
To get started, here are a few articles on HyperV by Janique Carbone on the VirtualizationAdmin.com web site that do a great job of explaining the basics of HyperV and some of the new HyperV features.
HyperV Integration Services: My interpretation is that this is effectively the replacement / upgrade for the old "Virtual Machine Additions" in Virtual Server and Virtual PC. You will definitely want to install these in your VMs (it makes the mouse 'usable', just like VM Additions).
http://www.virtualizationadmin.com/articles-tutorials/microsoft-hyper-v-articles/general/windows-server-2008-hyper-v-integration-services.html
Understanding and Using HyperV Snapshots: I still have alot of questions about HyperV snapshots, but so far, it seems like an amazing feature that works great.
http://www.virtualizationadmin.com/articles-tutorials/microsoft-hyper-v-articles/general/understanding-microsoft-windows-server-2008-hyper-v-snapshots.html
And here is a nice video she created showing how simple it is to create and use snapshots:
http://www.virtualizationadmin.com/articles-tutorials/videos/microsoft-hyper-v-articles/general/using-hyper-v-snapshots.html
What is pretty wild to me is how HyperV maintains 'branches' of snapshots, so it seems you can revert to a prior snapshot, then create new snapshots to create a new 'branch'. This is a pretty powerful feature, especially considering it only takes seconds to create or apply a snapshot.
But the caveat is that it can produce alot of permutations of your virtual server, so you'll have to learn how to manage those snapshots effectively so that they can be used in a meaningful fashion and not pile up over time to the point where they make a mess.
One small recommendation I have so far for taking snapshots is to try and log out of all sessions on the server before taking a snapshot. If you don't logout, your remote desktop connections will be saved as a disconnected sessions that you'll end up having to clear if you ever revert.
I've finally built a new Windows 2008 Server, so I'm going to be transitioning my virtual machines over to HyperV, which is the replacement for Virtual Server 2005. Because of the memory requirements of virtual machines, I've installed the 64-bit (x64) version of Server 2008, which allows me to get past the 3GB / 4GB limit of 32-bit windows. The desktop machine I used for Server 2008 only supports up to 8GB of RAM, but I figure that should allow me to pretty comfortably host up to 4 virtual servers (typically using 1.5GB of memory each)--more than I need running at any one time.
I'm still learning about HyperV and how it is different from Virtual Server 2005, and in which ways it is the same, so for the time being I'll offer some links to articles that found helpful.
My focus is not just learning and understanding the HyperV features, but also understanding how to perform equivalent maintenance operations and processes in HyperV to replace the tricks I learned with Virtual Server 2005. For instance, what is the best way to backup a HyperV image? The files that are created by HyperV are slightly different than VS 2005, so I need to learn what to backup, when, and how (i.e. Can I write a script to pause the VM and run a backup like I did with VS 2005?). Also, can I migrate an existing VS 2005 VHD file to HyperV (the answer is yes), and if so, what steps do I need to complete to get all of the server features working properly in HyperV? (still working on this one...)
To get started, here are a few articles on HyperV by Janique Carbone on the VirtualizationAdmin.com web site that do a great job of explaining the basics of HyperV and some of the new HyperV features.
HyperV Integration Services: My interpretation is that this is effectively the replacement / upgrade for the old "Virtual Machine Additions" in Virtual Server and Virtual PC. You will definitely want to install these in your VMs (it makes the mouse 'usable', just like VM Additions).
http://www.virtualizationadmin.com/articles-tutorials/microsoft-hyper-v-articles/general/windows-server-2008-hyper-v-integration-services.html
Understanding and Using HyperV Snapshots: I still have alot of questions about HyperV snapshots, but so far, it seems like an amazing feature that works great.
http://www.virtualizationadmin.com/articles-tutorials/microsoft-hyper-v-articles/general/understanding-microsoft-windows-server-2008-hyper-v-snapshots.html
And here is a nice video she created showing how simple it is to create and use snapshots:
http://www.virtualizationadmin.com/articles-tutorials/videos/microsoft-hyper-v-articles/general/using-hyper-v-snapshots.html
What is pretty wild to me is how HyperV maintains 'branches' of snapshots, so it seems you can revert to a prior snapshot, then create new snapshots to create a new 'branch'. This is a pretty powerful feature, especially considering it only takes seconds to create or apply a snapshot.
But the caveat is that it can produce alot of permutations of your virtual server, so you'll have to learn how to manage those snapshots effectively so that they can be used in a meaningful fashion and not pile up over time to the point where they make a mess.
One small recommendation I have so far for taking snapshots is to try and log out of all sessions on the server before taking a snapshot. If you don't logout, your remote desktop connections will be saved as a disconnected sessions that you'll end up having to clear if you ever revert.
Subscribe to:
Posts (Atom)