In the a previous version of OGG (220.127.116.11.0) we could configure the initial replicat as follows to process a trail created by an initial extract:
The replicat when started would then process the following files in the directory dirdat/E_LD_op:
However in OGG 12c (18.104.22.168.0) I receive the error message when starting the replicat: ERROR OGG-00045 Trails cannot be used with SPECIALRUN.
The solution appears to be to use the parameter EXTFILE and specify each individual file. However, as we promote code between different environments the amount of data will differ and thus number of files will change. The replicat process abends if a file is missing so it is not just a matter of specifying a potential maximum number of files.
Any ideas how I can get around this problem will be gratefully received.
Not sure if this meets the requirements of your use-case or not... but you can use "ExtFile dirdat/E_LD_op/LD_I MaxFiles 9999 [MegaBytes 2048]" which in effect creates a series of files just like trails . The "MaxFiles" isn't used in this case to limit the number of files created, but rather to force multiple numbered files to be created, so just pick an arbitrarily high number -- and make "megabytes" sufficiently large.
Message was edited by: MikeN -- Fixed "ExtTrail" => "ExtFile"
Thanks Mike. Unfortunately the parameters MAXFILES and MEGABYTES are only valid for the extract process and I will need to reference explicitly each file that the extract produces in the replicat parameter file.
You're right, that's not going to help... Sorry, I completely glossed over that you were using "extTrail"/"extFile" in a replicat parameter file, and not a pump (extract). In one last attempt to be useful (though, I haven't tried any of these, since I haven't done a "SpecialRun" since GG v9.5), you could use a variable in the prm file, so that you can script your replicat runs; e.g.,
That is literally "?" followed by any variable placeholder, e.g,. "myfile". So then, when you run:
bash$ export myfile=path/to/file000001
bash$ ./replicat paramfile dirprm/r_ld_i.prm reportfile dirrpt/out.rpt
The variable "?myfile" should be replaced by the env var $myfile... (That's how it works for normal replicat's & extract's, anyway.) Even better,
bash$ for f in path/to/file*
export myfile=$f # just being explicit here; could simply do "for myfile in ..."
./replicat pf dirprm/r_ld_i.prm rf dirrpt/$(basename $myfile).rpt
One one last idea (also not attempted!) is to see if you can go ahead and run "add replicat...extFile path/to/file" in ggsci, and then "start" it in ggsci, instead of running replicat from the shell and also remove the extFile parameter in the prm file. I'm not sure if that will work or not (I kind of recall that being an option at one point, but like I said, I haven't done something like this in a really long time.)
Message was edited by: MikeN
Message was edited by: MikeN