DB query issue

[eyetry]

Ok, I run a series of maps that load some tables, yada, yada, yaday....

The second to last map updates a table successfully. The last map reads the table updated by the last map and build a report, no variables and pointing to an mdq file.

As part of the chain of maps executed by the control map the last map always fails. If I run the map outside the control map it works. If I comment out all control map map executions but the last one it also works. Seems like theres a conflict. any thoughts?

6.7.1
windows

 

[janhes]

If you're running all the maps from a control map, has the update map committed the data at the end of it's run? What is the last map failing with?

 

[eyetry]

It fails with '8' invalid data. It's invalid because at least one row is required but the select returns 0.

I wondered about the commit. So, I inserted another runmap without any db connections between the the update map and the select map. The added just reads a 5mb file. Same results.

 

[eyetry]

The other thing I did to test this was to write the runmap rule as follows

seriestotext(dblookup("select * from table", " -mdq name.mdq -dn dnmame")) + "<nl>now run map<nl>" + run("map", "-ae=logfilename")

Which returned all of the data from the query + the comment + the return code, as shown below. So it appears the DB is avail for query.

data
data
data
now run map
8

 

[eyetry]

must have still been connected without the commit. instead of running a dummy map I executed this just before running the last map...

dblookup("quit", " -mdq name.mdq -dn dnmame"))

The result was that the last map ran without error.

think that's a little odd.

 

[janhes]

Could the problem be with "-ae=logfilename"? - path for the logfile not available?

 

[eyetry]

Thought that too. tried no logfile name and "-ae=path\logfilename". Same problem. Thought I had it working before I overrode the log file.

 

[snehareddy]

Hello all,

We have 5 instances of event servers running on HP-UNIX system.For each instance two processes should be shown – one Mercator process and one Java process.

It has been observed that for two instances multiple java processes are shown but there are no problems observed on Event server.Please let us know the reasons and remedy for why this multiple java processes are shown.

The version we are using is Mercator 6.7(306_1EM)

Thanks in advance!!