Deploying Peachtree integration shipping for Fedex

Posted at 11:25:55 PM in Hardware (8)

In November, when we initially started this process, we knew there was an import/export feature built-in to both Peachtree and Fedex where a daily batch could be exchanged, but this requires too many steps.  We wanted to implement integration to take advantage of several improvements.  These were:

  • Email notification to customers when a package was shipped.
  • Improved address accuracy (Sales Order number to Address instead of pulling the address based on a previous entry)
  • Tracking number and cost of shipment entered on the Sales Order (a process that was too cumbersome without integration)

Fedex offered a solution they came up with which used the ODBC driver supplied by Peachtree to connect to the database.  My client has two locations where this would need to be implemented.  One here in California which has been using Peachtree for over 10 years and one in Virginia which has been using Peachtree for only 2 years.  The California office discovered as they upgraded Peachtree every two years that the system was getting slower and slower.  They also discovered on one of the upgrades that they could no longer reasonably use the F3 find function in Peachtree.  Any query takes between 15 and 20 minutes to start showing data.  We assumed this was because of the large amount of data we have, but I just discovered that this same function also performs poorly in the Virginia office. 

The integration that Fedex offered was extremely slow.  It took about 15 to 20 seconds to pull up one ship to address for a shipment.  This seemed acceptable since it would take a lot more time to enter the address and then put the shipment info back into Peachtree, however; we discovered that all of the workstations connecting to Peachtree would freeze for that 15 to 20 seconds which was not acceptable.  This same issue appeared in both the CA office and in the VA office regardless of the amount of data.

The first solution indicated was Shipgear but Fedex recommended another solution Shiprush which they can "give tech dollars to buy."  I will not be able to discuss the Shipgear installation unless we run into problems with Shiprush.  If you look at the Fedex web site, you can see that V-Technologies has not be shoved off the plate by Shiprush.  I will document the Shiprush installation

2/22/2010

We were not able to use Shiprush.  I’ll explain the issue later.  We discovered that the Fedex “tech dollars” could be used to purchase Shipgear.  V-technologies is listed on the contract which opened the door for us to investigate using the original suggestion.  

At first glance, it appears that Shipgear  is primarily designed for UPS.  For example: if you go to the tech support page where you can submit a ticket, they only request the UPS meter number.  I suppose if you are using Fedex, that you put in the Fedex number and hope there is enough of a difference between the system numbering that they can't get  you confused with a UPS customer.  When calling tech support, they also ask for your UPS meter number.  But I found the software to be an excellent integration tool between Peachtree and Fedex FSM.

The workstation where Shipgear is to be installed requires Fedex FSM to be installed also without FXI (which is a integration package provided by Fedex).  The workstation does not require Peachtree to be installed (this works great if the Shipper normally doesn’t do anything in Peachtree).  However, the workstation does need to have access to the Peachtree data files.

To process a shipment after all of the software is installed, the shipper simply clears the Fedex screen and a Shipgear input field pops up over the FSM software for the shipper to enter the Sales Order number.  Shipgear can also be configured to pull the address info from the customer record or from the invoice.  Once the Sales order number is entered, Shipgear searches Peachtree for the address info and adds it to FSM’s ship screen.  From there, the shipper fills out and prints labels just like they normally would before Shipgear was installed.  Once the ship or print label button is pressed, Shipgear enteres the freight amount back into the Sales Order and updates the tracking info in either the notes for the invoice or the internal notes.  The FSM screen is cleared and the Shipgear Sales Order entry field is again displayed. 

A major benefit to keeping the FSM system is that all of the data from any previous installation can be carried over to the new system.  Our system has several package sizes setup so all the shipper has to do is select the package type and the dimensions are already entered.   We were able to backup the data from the old installation and restore it to the new FSM system and almost be 100% ready to ship.  I did need to get Fedex to dial in as there were some Admin functions that still needed to adjusted that did not come over in the restore.  And once you leave the FSM system, you cannot get any support from Fedex.

Issues with Shiprush:

  • The biggest problem we had with Shiprush is it's lack of ability to process multiple packages on a single order.  Every package needed to be entered separately. Between entries, Shiprush would wipe out the address, so pulling the address from Peachtree seemed to be defeated if you have multiple packages that need to be shipped.  This company ships several packages with every order, so this became a problem.  In Fedex, you can specify the dimensions and weight and click a button and type in the number of package that have the same info.  Then FSM produces a label for each box. 
  • A second issue was Shiprush uses web based tracking numbers.  These cannot be tracked using the regular tracking channels on the Fedex web site.  In order to track these orders, you need to install insight from Fedex, and then you do not have the ability to track by Sales Order number.  
  • Shiprush also has no facility to add other tracking information to the Fedex label / shipment.  It only puts in the Sales Order number.  It’s pretty common knowledge that the customer doesn’t need our Sales Order number.  They need their PO number.  The PO number, once entered was also wiped out in multiple package shipments.

Issues with Shipgear:

  • Shipgear runs several programs on the primary PC.  Among them are a web server, the program itself and the database server.  These programs demand a huge amount of memory.  Installing them with FSM consumes up to 800Meg of RAM which caused us some issues when testing the addresses.  When the data entry field pops up, if you click the button, it gives you options to pull a list of Sales Orders for a Customer or for a period.  These did not work.  When I called Shipgear for my “one” free tech support, they couldn’t figure it out and  I lost my one free tech support call.  The software was still usable because look ups on the Sales Order number did work though the tech support guy complained that our PC was slow and it should take as long as it did to pull up an address.
  • The system has been installed and running for over a week now and we are trying to iron out the freight cost formulas.  One of the benefits of using this integration software was to capture the freight costs.  We use FSM to increase our cost of shipping by 10% and Shipgear is to add a handling fee based on where it’s shipped and the overall sales order total.  We have not be able to verify that the freight is calculated correctly, but I think that’s attributed to the manual processes that were in place before.
  • There are not enough fields from Peachtree that would make ease of freight calculation easier.  For one, our Web orders already have freight and I don’t want to over write it.  One is if there is an amount in the freight already, I want to handle it differently.  I cannot access that field from Shipgear.

Benefits of Shipgear:

  • Shipgear has not caused any interference with any other workstation during shipping process. 
  • Orders can be tracked through normal Fedex channels.
  • No cut and pasting required to get data from Shipgear into Peachtree.
  • The web server running on the main PC allows users in the office to connect via their web browser to check on the status of shipping.  However, I do not know if the delivery confirmation can be checked on that web server. 

 

Written by Leonard Rogers on Monday, January 31, 2011 | Comments (0)

Cargowise EDI setup for SQL database at remote location

Posted at 10:18:34 PM in Installations (47)

The database engine behind Cargowise is MS SQL.  We installed the 2005 Express package that came with Cargowise's installation package.  The ideal software installation wants everything on the same network.  This becomes important when setting up file locations in the admin section of Cargowise.  These file locations are processed by the Controller (which is a PC that communicates with the SQL server to run jobs such as FTP, email and print jobs).  The file locations are always relative to the computer that runs the job.  In the case of FTP, it is relative to the Controller.  However, in the case of SQL database backups, it is relative to the SQL Server.  In our installation this becomes a problem, since our installation houses the SQL database off site.  We set the backup path originally on the P drive.  The P drive is common to the workstations and Controller, but not to the SQL Server so the backups wouldn't run.  We had to change that drive to one that the SQL Server could see.  This naturally also creates a problem when upgrading the software.

Once we had all of the workstations installed and the server installed at the remote location, we couldn't get the two to talk.  Cargowise's techical support was very helpful here.  To access the server, we had to run a SQL helper utility called cliconfg.exe.  It ships with Windows XP, I’m not sure about Vista or Windows 7.  This application allows you to set an alias that will provide the IP address to the server.  (see: http://blogs.msdn.com/b/sql_protocols/archive/2007/01/07/connection-alias.aspx).  This allowed us to give an alias name to the server.  But it would not make the connection unless the alias name also contained the instance name as well.  In this case it was server/edienterprise.  Then we were able to make the connection and finish the setup.

Upgrades need to be performed on the SQL server first.  The workstation client is also installed there to perform admin functions like upgrades.  Once the server is upgraded, we copy the shared upgrade folder to a common location on the local network and then update the clients and controller from that shared location.

Written by Leonard Rogers on Monday, January 31, 2011 | Comments (0)

Print Server Properties

Posted at 1:32:17 PM in Other (7)

In Microsoft Windows 7, the print server properties option doesn't show up on the options bar until you've selected a printer.  To access that option, you have to be in devices and printers, then highlight any printer and additional options will appear.

Written by Leonard Rogers on Saturday, January 29, 2011 | Comments (0)

win 7 64 bit HP laserjet 1020 install

Posted at 1:13:12 PM in Hardware (8)

What a pain.  If you go to the HP web site and enter the HP Laserjet 1020 as your printer and then pick Windows 7 64-bit for the OS, it gives you drivers that it claims are compatible with Windows 7 64-bit.  Not so.  

The drivers are actually for Vista 64.  The z-7 zip file complains that it won't run on a windows 7 64 bit machine and to contact the vendor.  So you end up having to use winrar to extract the files.  Then when you run setup.exe, you get a similar message that says it's not compatible with Windows 7 64-bit.  Very frustrating, since HP offered the driver as a solution for Windows 7 64-bit.  And yes, I submitted feedback on their web site. 

Finally, I used these instructions on HP's forums that helped a lot.  See: http://h30434.www3.hp.com/t5/Printer-All-in-One-Install-and/install-HP-Laserjet-1020-printer-Driver-windows-7/m-p/145921.

Be sure to check the video links for a better explanation.  The opening instructions has a lot of assumptions.  I did my install all from devices and printers and there was no install driver from specified location option.  I used the troubleshooting option which worked to reinstall the drivers.  On my system, it had been working.  I don't know if the customer moved the usb cable to another port or what reason it suddenly disconnected itself from the system.  I had a printer driver that indicated it was working, but wasn't and a HP 1020 unidentified item on the devices and printers screen.

I removed the HP 1020 printer and went into print server properties and removed the driver as well.  Then I right clicked on the unidentified icon and selected troubleshoot.  The trouble shooter found the driver and installed it again and it worked.

 

 

Written by Leonard Rogers on Saturday, January 29, 2011 | Comments (1)

Apache server not responding on any IP or port

Posted at 3:14:18 PM in Installations (47)

After installing a minimum install Fedora 14 text mode, I yum'd httpd and started the service.  I could not get it to respond on any IP address.  I ran

netstat -tun | grep 80

and found that httpd was running and listening on port 80.  After searching around, I found that iptables and ip6tables had been installed as part of the minimum install package and I'm sure they were setup as a workstation which wouldn't have a web server.  I ran  

service iptables stop

and checked the web server again and it worked.

Now to figure out how to configure iptables in text mode. 

Written by Leonard Rogers on Wednesday, January 26, 2011 | Comments (0)

Fedora 14 Install experiences

Posted at 2:44:54 PM in Installations (47)

I have been trying to get Fedora 14 to install over the network.  Everything I did failed all around the missing repomd.xml.  The only installation that worked was using the DVD.  I'm updating an old server that doesn't have much memory which are all problems and I didn't want the graphics client installed.  I did the network install because it allowed the use of a kickstart file.  Once I started using the kickstart file, it refused to look back at the CD for the repository.

To get a text based server installed, I used the second menu option from the install screen, then pressed tab to add options and typed text at the end of that line and pressed enter.  The install was minimal.  It did not install any graphic desktop files.  I believe the total package count was 195.

Once the setup was done, I was able to boot and login, but had no network connection.  NetworkManager was not installed and setup use to provide a text menu to configure the network but neither of those are installed.  So how to get my network card working???

After researching, I found that I had to edit the ifcfg-ethx files.  I have two cards and from all the network installs I was trying to do, I was able to easily determine which one was attached to my router and could be assigned a dynamic IP address.  The other network card requires a static IP address which I will figure out how to setup first.

First thing, I needed an editor.  vi apparently comes with all of the Linux / Unix installs, but my favorite is nano.

My dynamic IP address would be assigned on eth1, so I used vi to edit ifcfg-eth1 and set the type to BOOTPROTO=dhcp and changed onboot=yes.  There was a line that wasn't in any of the example files so I removed it.  I appeared to be related to NetworkManager even though that wasn't installed on the minimal text install.

The resulting ifcfg-eth1 file looked like this:

DEVICE=eth1
BOOTPROTO=dhcp
ONBOOT=yes
TYPE=Ethernet

I restarted the network service: service network restart 
I also used dhclient to cause the dhcp service to acquire an address.  Without that, we could not get the network card to acquire an IP address, so we weren't on the network.

eth0 needed to be static, but I needed to start installing yum updates.  After I finished the updates, I changed eth1 to a static address also.  The above ifcfg file stayed pretty much the same it is listed below:

DEVICE=eth1
BOOTPROTO=static
ONBOOT=yes
TYPE=Ethernet
IPADDR=192.168.101.2
NETMASK=255.255.255.0
GATEWAY=192.168.101.1


 

 

Written by Leonard Rogers on Friday, January 21, 2011 | Comments (0)

repomd.xml repository installation file used by yum

Posted at 11:01:36 AM in Installations (47)

Many Linux installations (my favorite is Fedora though I'm not sure why) use yum to update their program files and libraries.  During recent installation attempts with Fedora 14, I kept getting error messages saying repomd.xml could not be found which is a very important file used by yum to perform it's upgrades.  I discovered that this file is normally found in a directory called repodata off the base directory provided.  So if you are trying to provide the location for this file and it is located on an HTTP server or FTP server, you would specify the base url as http://host/os where the os directory on the host has a folder named repodata which contains repomd.xml

I will be making several entries regarding my Fedora 14 install.  I've been at it for over 24 hours and it has been a terrifically difficult process.

Written by Leonard Rogers on Friday, January 21, 2011 | Comments (0)

First Apple iPad Install

Posted at 10:19:07 PM in Hardware (8)

The new Apple iPad is the first Apple appliance that I had considered buying and using for myself.  I use ebooks a lot.  I like reading and my cell phone is just too small.  It's an added plus to have email and web browsing available and an abundant number of apps available for download.

I've had to setup the email on several iPads in the past couple of months. I was pretty happy with the interface and general feel.  It had the feel of an over-sized iPhone; an easy to use and navigate touch screen.  Most of the menus are exactly the same as the iPhone.  I recommended the iPad to one of my friends instead of one of the other ereaders because of the added features.  He was concerned that he'd have to buy a data service such as AT&T to use the iPad which I cleared up.  The iPad has WiFi capability, so if you have wireless internet, you can use the iPad's email and web apps and you don't need the internet once you've downloaded the ebooks.  He purchased the iPad and offered me a chance to setup my first iPad from the ground up.  It's still a good app for him, but I have some reservations about it's utility for me.

I had an issue with the lack of a java enabled browser.  As far as I know, the browser brand that they use in the iPad (Safari) can normally execute java apps, so it appears that they intentionally gimped the browser.  The first thing I wanted to check after connecting to the internet was the internet speed at speedtest.net.  Rather than bring up the web page, the browser shut down and pushed me over to the app store where i could download a speedtest.net app.  At the time, I couldn't download the app because I didn't know my apple id.  It was free, but I didn't want the bloat for a one time test.  I doubt my friend would have ever used the app.  I tried several more times and tried different ways to get to speedtest.net, but ended with the same results.  It was driving me bananas.  So I used another web site that does speed tests and found that the java app wouldn't install.  Then I tried to install java from java.com and that didn't install.  The error message I got was Operating System unsupported.

I do a lot of remote work in my office, so I tried my gotomypc account and that wouldn't work either.  This alone would prevent me from considering the iPad, but this is not an issue for other users.  We finally reset our apple id which was setup at the store and my friend couldn't remember the password.  The apple id is required to download music from itunes and download apps.

What really ended it for me about the iPad is when we wanted to download a free app and we had to enter credit card info.  It wouldn't let us until we stored a valid credit card.  "store" being the operative word.  Every indication was that our credit card would be stored for future purchases.  The item was free.  This would not have bothered me had their been a charge, but the item was free.  I'd rather put in the credit card info every time I buy something or even go with the itunes structure and put a certain amount of cash on account and pay using that account until it's empty then add more cash later.  But this seems to store the credit card info and when you do buy something that costs money (which sounds dumb, but in the apple store, you can buy something that's free, so...), they just use that card unless it doesn't clear.

Anyway, I didn't get to try buying something that actually had a price to see if it'd submit the credit card info without requiring information to complete the transaction other than saying yes.  However, I got immediate nightmares of children playing with the iPad running amuk buying everything that peaked their curiosity and then I'd have to figure out how to get that money back.  If that is the case, then I'd wrap up the appliance and return it asking them to cancel my apple id, if that could be done.

I like the iPad.  It's still a nice appliance, but gimping the internet functionality and storing credit card info for items that are free from their app store... that's just too much.  I think the iPad is going to mature like the iPhone did and prehaps in a couple of years, they'll have an appliance I can use.  Until then, I'm going to buy a laptop.  It's cheaper anyway.

 

Written by Leonard Rogers on Monday, January 17, 2011 | Comments (0)

Follow up on Web Master tools

Posted at 9:53:55 PM in Web (9)

I set up my site on 3 major search engine Web Master Tools sites: Google, Bing and Yahoo.  I was surprised to see that Yahoo was submitting my sitemap information to Microsoft.  Apparently, they are collaborating.  I placed my information on those search engines back in 2009.  Yahoo and Bing have not checked or re indexed the sitemap since I set it up originally.  So much for submitting a sitemap.

Google has been checking the sitemap.  They have a page that shows the errors in the site crawl and the date they crawled it.  Sometimes, I wish there was a way to get errors off the crawl list when you've fixed them, but it eventually clears up. 

Basically, I find the web master tools for both Bing and Yahoo are a waste of time.  At least Google appears to be doing something with the tools, while their competitors just collect the initial data and then let it sit.  They must be short of man power or resources.  The web master tools are collecting dust at Yahoo and Bing, or should I say, Microsoft.

 

Written by Leonard Rogers on Monday, January 17, 2011 | Comments (0)

IPhone on Verizon's network

Posted at 12:21:05 PM in Vendors (35)

Reading todays business section of the LA Times prompted me to offer my opinion on the iPhone saga as well.  The article "IPhone may put strain on Verizon's network" stated that a large number of iPhone users would be clamoring to get on Verizon's network and that may bog it down.  Since the number of iPhones sold in 2010 was around 12 million, they expect that Verizon will take about half of that market share in 2011 if they add the phone to their list. (The article didn't actually say Verizon was going to offer the iPhone this year, however; if it does offer it, it will show up on the array of phones on Tuesday. I don't know if that's today or next week Tuesday.)  Speculating that half the existing iPhone users (estimated at about 73 million) may switch their service, Verizon could stand to take a big hit on it's network performance almost overnight.

Such an exodus is quite possible.  The article indicated a general dissatisfaction with AT&T by iPhone users stating they complained about "dropped calls and slow performance."  I was an AT&T customer.  I dropped them because of those same reasons.  I was so dissatisfied with their service that I wouldn't switch back even for the iPhone.  When I was a customer, I could watch the signal snap from one tower to another and still drop calls in the switch.  Most often, the dropped calls were random.  After I dropped AT&T and switched to a local company, AT&T acquired that company and within 2 months the service was back to the same problem.  I speculated that AT&T bought my cell company so they could dump their current customer bloat onto additional towers giving their current customers a little better service.  In reality, my cell phone company may have been using AT&T's cellular towers all along.  If the later is the case, then problem isn't with the equipment or the tower placement, but the problem would be in the management of that equipment.

AT&T claimed that it's network was 60% faster than Verizon's network.  I'm not sure what difference that makes if you can't stay on the network.  They could also make that claim because nothing in the claim says has access to that bandwidth. I'm sure anyone currently on AT&T might be wondering why they don't have access to it.  However, the claim probably is implying that if the service is slow now, those that switch will experience more than 50% poorer performance on Verizon's network.  My argument is if you keep dropping calls, you can't very well stay online.  That would certainly result in poor performance.

Unfortunately, I had a chance to use the Android phones with the Google apps and I'm quite happy with my phone, so I won't be one who will switch.  But if you can't stay connected to the provider's network, then perhaps it's time to look for a new network, now that you can.  Good luck AT&T users.

Other links of interest:

Written by Leonard Rogers on Tuesday, January 11, 2011 | Comments (0)

Cargowise EDI

Posted at 5:02:41 PM in Vendors (35)

Cargowise EDI is vertical market software designed to help freight forwarders to track their shipments, manage customs documentation and notify their customers of the status of those shipments. The software can be purchased as a standalone application (though it requires a constant connection to Cargowise servers which constantly check the user license and modules used) or as SaaS (Software as a Service)  The standalone version allows you to host the database on your own servers with the requirement that you always maintain your server, SQL database engine, web server and all the workstation with up to the minute Microsoft releases. SaaS provides all of these upgraded functions for you and all you have to maintain is the workstations. Standalone versions have a monthly maintenance fee and an annual contract payment which can be paid monthly. The connection requirement to Cargowise servers allows them to stop the software from working if these fees are not maintained. The data is locked behind a password that Cargowise will not give you and they go out of their way to prevent you from accessing the data directly. Of course, in an SaaS environment, you don't have access to the data anyway. It makes since that if you don't have an IT department, the best way to go is with the SaaS solution. 

Note: The customer that I work with that uses Cargowise EDI has complained that every upgrade has lost data, stop routine reporting or completely removed functions that forced them to find new ways to accomplish the same task. The frustration they constantly complain about is the lack of support from Cargowise. They break something in the upgrade then tell you it's not their fault.
 
Here are a  few annoying things about Cargowise EDI. These are all complaints from the customer who actually uses Cargowise. Every new software installation, especially in the vertical market is going to be awkward in the beginning. The only people the software is intuitive too is the programmers, who you never get to talk to. These programmers are from Australia and the language is a huge barrier. Terminology and translation is the first step to bridging this gap and Cargowise seems to make little effort to addressing this. The next step is translating the customers way of doing business into the way Cargowise does the same thing. 

1. Training is expensive and unproductive.  The trainer cannot or will not answer questions outside the curriculum.  First and foremost, training should address how to get into the software, how to process an order and how to get out of the software.  Everything else should be an appendage to that and should point to it.  For example, in order to get into the software, you have to set up a user. In order to set up a user, you have to specify what role the users will have. Each role grants access to certain portions of the software which means we have to define each area of the software. Cargowise's instructors and instructor guide completely miss this point. The customer feels he is learning something that he will never use.  For example, my customer didn't want to do accounting in Cargowise, yet we had to learn accounting without an explanation as to why.  After 2 weeks of training, we still couldn't process an order.  That was a problem.  We ended up paying someone who had experience using Cargowise to come in and teach us how to use Cargowise and customize it to our needs.

2.  SQL administrator lock down security.  Once Cargowise is installed, they lock down administrator functions and change the sa password.  This creates a problem when trying to administer the database.  You have to contact them for support even for the most mundane admin tasks.  

3.  The web interface is .net and extremely slow.  Every visitor must wait at least 30 to 60 seconds for the page to load the applet(s) on the first visit.  (Note: SaaS doesn't seem to have this problem, which means server specs probably has a lot to do it). If you can get by that, the response is fairly quick.  But data extraction in both the workstation application and the web site is also extremely slow.  There are a lot of tabs and each tab has a lot of information and drop down menus.  As the application is practically all data driven, every time you load a page, the SQL server queries all of the look up tables, drop down lists and record information for all of the tabs being displayed.  That's a lot of information to pull across the internet.  If we had hosted all on the same network, everything would respond quicker.

4.  This isn't really a problem, but it is annoying... They like to keep up with Microsoft's new updates.  We bought the software with Windows 2000 workstations and were told that the OS would work fine.  But the very next upgrade, we were told that we couldn't unless we had Windows XP installed, so we had to upgrade all our workstations.  We installed on SQL 2005.  Now, we're behind held off our 2nd upgrade because we don't have SQL 2008 installed.  This kind of requirement is just too much for the average office manager or freight forwarder to have to deal with.  Although there are many savvy users out there today, shouldn't freight forwarders be able to concentrate on freight forwarding?

Cargowise is an excellent application for freight forwarding.  We researched several packages and found that this one met our needs best.  The installation and support seem to work best if everything is housed on the same network and a web port opened to the server.  Hardware and software upgrading at the whim of Microsoft's upgrades can sometimes be a real pain.  The tech support is quick to respond to questions.  I am not addressing the functionality of Cargowise.  It works.  My intent was to discuss the technical aspects of Cargowise as I saw them during installation.  I will be writing other articles as issues arise during the use of Cargowise.  At the time of this writing, we've been using Cargowise for a little over 2 years.  

My next article will be about some of the issues of installing Cargowise in our weird configuration and how we have to deal with upgrades.

Written by Leonard Rogers on Monday, January 10, 2011 | Comments (0)

Peachtree reports and forms missing

Posted at 1:06:43 PM in Recovery (43)

Recently I had an issue with Peachtree.  All the reports and forms disappeared based on the customer's report.  I connected to their server and used Peachtree to check the symptoms.  Every report in the reports window was missing.  No custom reports and no standard reports. 

After searching the web, this link was most helpful.  StephenC's entry about the difference between the custom reports and the standard reports pointed me in the right direction.  In my case, it was both report groups and I didn't know they'd be in different data files and on top of that that the files would be in different directories.  It would make sense that the standard files would be in a common directory to all of the companies.  At this particular location, we make a backup every year-end and then create a new company from the backup so they can access old info.  I had the controller open an old company and sure enough, the custom reports were there, but not the standard reports.

RPTDATAI.DAT in the main datapath for peachree was truncated to 14K.  The standard reports file in the backup was 2900K.  I was able to restore rptdatai.dat in the main datapath and restore rptdata.dat in the company directory and all the reports and forms showed up again.  There is no indication what caused this.  None of the users reported any odd activity when closing down peachtree or during it's use.

This installation uses Quantum 2009 on a Windows 2000 platform using active directory services for authentication.  All the workstations are WinXP Pro SP3.

Written by Leonard Rogers on Monday, January 10, 2011 | Comments (0)

Rant on Antivirus Software

Posted at 4:29:45 PM in Recovery (43)

I got a call today from a customer who uses Outlook 2007 to access his Gmail account which is branded to their domain name.  He said that for some reason this morning, his Outlook stopped working.  The error message said it was unable to connect to the server.

I checked all the settings in Outlook and made one change to the outbound port number but that didn't fix the problem.  Then I checked Google's Gmail settings and everything there was fine.  I then took a look at the antivirus they were using, Trend micro, and started turning off services.  I have had this kind of problem before with Norton and never could figure out how to get Norton to behave.  My theory is that a virus infected email was blocked by the anti-virus software but it blocked all of the rest of the emails like plugging up a hole.  Unfortunately, it never unplugs the hole again.  For Norton, I would have the customer uninstall the it and then reinstall.  It was amazing that email worked fine after the Norton Antivirus software was reinstalled, so apparently there wasn't a virus infected email server after all.

However, with Trend micro, we found the problem in the Internet protection section.  When we disabled the web protection, the software warned that we should restart the browser. From the description, this didn't appear to be the problem area, but with that disabled, email started coming through.  When I enabled it again, the email stopped.  The customer mentioned that the antivirus gave an indication there was a problem earlier but didn't know what the message meant and just accepted the default answer. I suspect that a warning about a port opening may have popped up asking him what he wanted to do and he blocked it.  It's easy to block things in antivirus software applications, but it's not so easy to undo it.

Now comes the part where I make my recommendations about anti-virus software and what it should and shouldn't do.  I've re-written this several times because I find it very self-incriminating.

I believe antivirus companies should work on protecting you from virus' and leave the other items like firewall and spam protection to other companies. Microsoft's Windows Defender is one of those that should be left to other companies develop. Microsoft installed it on Vista and Windows 7 and provided no way to uninstall it and it complains when it's turned off.  For spyware, I like to use my own choice of software. Microsoft's apparent motto is "protect the user, don't let them open anything." There was a lot of grief because Outlook simply blocked certain attachments from coming through in email.  The attachment was there, but it wouldn't allow anyone to open it (without a hack), you couldn't even see it.

Antivirus software vendors should not add firewall services to their software, but they do.  In fact, I don't think you can buy any antivirus software without the firewall option being available.  I realize that virus' can spread through a network and a firewall can help prevent it.  But it isn't virus protection and the email example above is typical.  The user gets blocked and no one, not even the vendor's support persons know how to unblock it.

Now we have a whole array of protection being offered by AV companies, i.e. anti-phishing and identity theft protection among them.  I think most of them are gimmicks to show how much better they are than the other product.  If Norton adds something, then all the other AV companies have to add it too. It doesn't matter if it actually works. Some vendor's alert you when another antivirus is already installed. Trend micro won't allow you to install if the other antivirus software was there. They alerted me to having the free version of Malwarebytes Anti-malware software, which doesn't run real time and shouldn't have been a problem. I called tech support at Trend micro and found this was by design. So, I uninstalled Malwarebytes software and reinstalled it after I installed Trend micro.

Most of these antivirus softwares eat CPU. With more features, they eat more CPU. In almost all of my installations, I move the virus scanning for email, anti-phishing and spam protection to a firewall appliance.   One server does most of the work protecting the network and the PCs only worry about virus'.  For the firewall appliance protection, I use IPCOP with the COPFILTER addon.  I use the free version of CLAMAV and it seems to work very well.  IPCOP developers are now coming out with version 2 of the server after many years of supporting version 1.4.3.  Copfilter will have to catch up.  Copfilter has been a great product.  They have a disclaimer that no antivirus will catch all virus' and that is true, but this package has worked very well for me.  

That's my rant on Antivirus softwares.

5/21/2013 - edit.

I'm now using IPCOP V2.06 and Copfilter 2.0.91beta3. Spamassassin seems to be working pretty good. I'm not happy with the 3rd party signatures (part of Copfilter's CLAMAV component). The idea behind the 3rd party signatures is to use CLAMAV to identify email patterns that are commonly used to phish or spamvertise .i.e One looks for emails that consist only of a image which is one way spammers circumvent spamassassin which is a word database... no words, no spam, but the 3rd party signatures component addresses this and many other patterns.

The reason I'm not happy with it is because the installation isn't consistent. I have two identical machines and one isn't catching any virus' and therefore no patterns and the other one is. I cannot upgrade the 3rd party signatures at all. Frustrating at this point.

Written by Leonard Rogers on Friday, January 7, 2011 | Comments (0)