Trying to do a simple query in my IDE against a perfectly fine FMP 16 database gave me this error today:
Simple two-table join that used to work.
--
I have a now long-standing bug report into FMI regarding JDBC and its non-standard dates and incorrect IDE date display (wrong date shown -- no word back in the "supported current version" with any updates).
Local.
Thanks,
any chance another process was touching the file at the time? Indexing, AV, backup,...?
Nope, I'm the only user of this computer.
To the posting below, there is no "JDBC Service", I'm just using the FMP JDBC driver and its API.
----
To debug situations like this .... Here's yet another example where FMP needs to have an external log like other products.
Maybe that's your problem
Even run locally, there should be an "fmxdbc_listener"
greg
> To the posting below, there is no "JDBC Service"
A one-table join does complete...after a minute.
Yes, the listener is loaded, but takes 99% of the CPU.
You know, there should be no way a user could generate errors like this.
I think the problem is that there are too many fields for FMP to handle in the SQL ResultSet, but the error message doesn't give you any clue.
FMP's JDBC driver is seriously deficient (wrong results shown, errors that appear to have nothing to do with the actual problem, many unresolved-but-reported bugs, ...) relative to other drivers I've been using since 2002.
Not trying to start a flame here, but as this deficiency has baffled me for years, what type of serious development product doesn't LOG issues and problems for analysis and debugging? (So, ... Constant posting of problems here on the forum and FMI's only actionable reply, that is never allowed by most companies is to ... "send us your solution." Uh, no, sorry, can't do that.
No LOG -> Problems remain...
There is a similar discussion here: http://mightydata.com/monitoring-the-filemaker-server-xdbc-process/
FileMaker is exponentially slower than a "real" SQL DB, so don't expect miracles
It seems the xDBC Listener was "written by a Third Party"
greg
> Not trying to start a flame here, but as this deficiency has baffled me for years
Not sure what you mean, mathematically (as in how you quantify that statement) as 'exponentially slower', but I get your gist.
"Written by a third party" ... begs the question: who is ultimately responsible?
FWIW, I have never (ever) seen a JDBC driver bug in MySQL, SQL Server, or Oracle -- three environments I use for most of my production work.
---
I appreciate your links but I need logs or actionable info so I can infer what's going on.
FMP gives me nothing in this regard.
Is that against a hosted file on FMS or a local file?