Our company uses approximately 135 remote databases deployed on Laptop
computers. Rarely does a week go by where we do not have at least one
or two assert errors and need to re extract the databases. But on
some weeks it can be as many as 8 or 9. Some of the assert message
errors are pasted below:

*** ERROR *** Assertion failed: 201502 (10.0.1.3887)
Inconsistent page modification counter value

*** ERROR *** Assertion failed: 201501 (10.0.1.3887)
Page 0x0:0x1870cf for requested record not a table page

*** ERROR *** Assertion failed: 200502 (10.0.1.3887)
Checksum failure on page 2065193

My question is this: Is there any way to find out what is causing
these to occur so frequently? I've delayed posting this question in
the hopes of finding a discernible pattern among the receives and I
really haven't been able to. The closest thing to a pattern is errors
during replication, when about half of the errors seem to occur. Our
mobile workforce receives via dbremote only once a day, and I see in
many (but by no means all) machines with these errors files the assert
will occur suddently while the machine is applying messages.

The remote machines are mostly Lenovo laptop machines, two and three
years old. We DID notice when we first got these machines that the
databases seemed less able to successfully recover from being shutdown
improperly than the previous generation of machines we had been
using. For example, if the database appeared to not be responding and
the user killed the application using task manager, the database will
try to recover when it is started again. We have observed that these
recoveries more often fail on the new machines.

I realize this is not much to go on for such a big issue, but I was
hoping to get some feedback from anyone who might have had similar
problems in the past and at least gain some insight on this issue.

Thank you

-Tim O'Connor