Wiki Home

Invalid Seek Offset Error 1103

Namespace: VB
I have been getting this error with an application I run to do some behind the scenes processing and it has been partnered with the "Windows - Delayed Write Failed" I am pretty convinced it is tied to network disconnectivity. The solution for me has been to reboot the machine and restart the background process.
I am looking for information about the possible causes of this error. We are running a VFP 8.0 application (SP1), with the main program executable and the data residing on a mapped network drive. The server is running MS Small Business Server 2003. The client computers are running Windows XP. After the FoxPro program is running for a while, the Invalid Seek Offset error will pop up. Of course, I cannot trap this error so I do not know what line of code may be triggering the error. I also have the program running on a Windows 2000 machine under the same conditions, but have yet to experience the error on this machine, so I am wondering if it is an XP thing. In addition, there has also been a "Windows - Delayed Write Failed" from the OS on one machine. Do these problems have anything to do with the fact that I am mapping to the executable rather than running a copy of the executable local on each client pc?
Part of the cause seems to be somewhat frequent disconnects from the network, which we are addressing, but is there a way to make the FoxPro app react to these problems in a "nicer" way? And, what is generally thought to be the cause of the Error 1103?
Thanks for the help.
We've seen this happen mostly with marginal or intermittent network connectivity. Don't have a good way to troubleshoot it beyond maybe checking the system's event log to see if there's anything logged about TCP/IP or NIC connectivity events. Good luck, this one can be a bear. - William Fields
ISO error means that there is index corruption, likely caused by those network disconnects. The operating system and/or network hardware is your primary suspect. VFP can't really be expected to compensate for hardware and OS-level services that are intermittent, short of a major re-write of your data access logic to open, write and immediate close every table. Calling in an experienced network engineer could speed the solution, or you can try isolate the problem to a single machine. -- tr
Sorry, but I don't think index corruption has anything to do with the ISO error, although the error may certainly cause corruption. Every time my users have hit the ISO, the immediate solution has been to just restart the application, and every user went right back to wherever they were when the error hit, with no problems. If there had been corruption, I would have expected at least some users to hit it again on the same records, until the tags had been rebuilt. -- Alan Rutledge
Over the years, we had seen this error and determined cause to be a memory conflict with printing systems from HP and a small number of other vendors. We fight this much less now with VFP9 SP2. VFP constantly polls the default windows printer whether you are in the act of printing or not, so there is no real natural indication it is the printer causing the problem. Try uninstalling HP printing systems or other specific vendor printing systems that exist and use only the equivalent driver for your printer from microsoft. - Stacy Violett
We have encountered this problem at one of our clients. We are currently writing a loader program that will maintain a copy of all APP files on their local hard drives (i.e. update them if newer version), and run them from there. This should improve performance anyway, but we wonder if it will eliminate the "invalid seek offset" issues at that site. I will publish our findings here in due course. Graeme Evans ( Mar 29, 2010.
We triggered Invalid Seek Offset Error 1103 when trying to process a 7GB file on NTFS hard drive using VFP 9 that had been installed on a FAT32 drive. After breaking the 7GB file out into 1GB files, our 1103 errors vanished. --- bjg
I am finding that we get this error even on the server. Possibly memory. seems to go away after reboot.
( Topic last updated: 2011.06.30 12:31:25 PM )