1 of 1 people found this helpful
check for DMP files in the FMS logs folder, also check the FMS stats: are any maxing out (assuming the default measuring interval is still set at 15 seconds; are any stats reporting numbers in the 15,000,000 range?)
Thanks! I will check.
Something is tying up the processor and makes FMS queue up the requests. Check the system processor stats to see what it is beyond FMS itself 9f anything (I doubt it but it is worth checking).
I don't see many "calls" being executed though.
Did you upgrade the drivers at the same time as switching to the new server? If so, see if you can switch back to the old drivers. ESS may not work well with drivers released after the ones that were required at the time FMS was released.
Same problem before driver update; I hoped the update was going to cure it, but it made no difference.
Sometimes it seems that I can scan through an ESS record set OK.
My system creates matching records in a FileMaker join table.
It appears when I come upon an ESS record that does NOT yet have a matching FileMaker record, this appears to be when the problem occurs.
I get a Dialog "Find in progress... Processing query"
Is it completely locking up? Or just taking a long time?
How big is the table? About the worst case scenario for a search is when there is no record to be found. It seems to be a bigger hit in SQL than in Filemaker, but there's a hit for sure. Have you checked your indexes for corruption?
There are some related tables involved. The primary MySQL table now has 337 records.
There may be a variable number of related ESS childe records.
The problem does appear to occur where there is a variation in the number of associated child records.
Check indexes for corruption.
Thanks for the suggestion.
But I'm not clear on which indexes. FileMaker indexes; or mySQL indexes? And if mySQL; check indexes how?
I will be going in to see the client again in an hour. This is a severe problem.
If a FileMaker Pro client is connected but "hangs" and you use server admin to try to disconnect the client, what may happen instead is the FileMaker Server shows as closed, and all files come up as not closed properly.
Still hoping to find further info on how to check indexes; or any other clues.
I will look forward to reading about what you discover as there are obviously situations where ESS is treated like a foreign object by some IT admins.
We encountered this when trying to get the ODBC driver connected to a new datacaentre. Fortunately I had a colleague with me who knew the expected behaviour, knew what what we should be seeing, and could be persistent with the IT admin. Had that person not been with me I could easily have been bamboozled by the geek speak.
All the best for 2014!!
In SQL Server and Oracle, you can simply drop and recreate the indexes, assuming you have DBA access. Not 100% sure about mySQL. Filemaker, you turn the indexing off I believe, then quit, restart and then re-enable the indexing. You may want to run the data utilities as well.
Are you saying you have more than 15M Filemaker records? One other thing to check.. Make sure the Actual ODBC driver on the Server has a full serial#. Otherwise it runs in demo mode, which lets you scroll 3 or 5 records if I recall and then stops. Also make sure it's not a standard desktop serial#, because that will limit the number of ODBC connections that can be made. Jonathan sells an "unlimited" license specifically for Server.
Let us know what else you find.
Here is some information I got from Jonathan Monroe of ActualTech.
Changing this setting seems to have helped:
I suspect that your MySQL connection is being dropped by the network - either because of inactivity, or a dropped connection by some piece of network equipment between the driver and the MySQL database. The "freezing" is occurring because the driver has not been notified that the connection has been dropped, and it is continuing to try to use the connection. The default timeout duration while attempting to use a connection that has been dropped is around 10 minutes.
We added a new option to the 3.0 driver that may improve your situation. If you have the latest driver installed, please press the Advanced Options button on the Connection panel of the driver's DSN setup assistant.
Try changing the value of the "Network timeout" field to something like 15 or 20. This should cause the driver to "fail quickly" if the connection to the MySQL server is dropped. So, instead of hanging for 10 minutes before getting the error message, it should only wait for 45 seconds (i.e. 3 retries x 15 seconds).
You shouldn't make the number *too* small, because it does need *some* time to be able to connect, especially if the MySQL server is busy and is located at a remote facility. But, hopefully you can find a compromise value that will allow the driver enough time to connect, but not wait an unnecessarily long amount of time to tell you when the connection has been dropped.
Please let me know how this works for you.
Lee: I'm not saying i have more than 15M records. One table has 1,400 records. The other table has 400 records.
Record counts are tiny and should present no problem at all.
I'm not sure where the 15M figure comes from or where you think I might have mentioned a number like that.
The server stats show delays of over 15M milliseconds. Maybe that's the point of confusion.
Sorry, reading too fast again.
Did the info. Jonathan provided make any difference?