> Hi Steve, > > Thanks for your reply. I am almost emancipated from the ipc system > except I use gptext to create a label. > I am experimenting gptext using DEVICE = VG | file.vgf. Since this also > uses queueing to communicate with the device driver, presumably "vg", I > would like something like gptext_vg. I don't see it in the makefile. > Before I invest alot of time in this, do you have any suggestions? > > Should I be using something other than gpmap and gptext? If this is my > only solution, is > > Thanks, > > Mark Hughes > Global Weather Dynamics, Inc. Mark, I've attached a modified Makefile for $GEMPAK/source/programs/gp/gptext The _vg looks just like the _nc options with th VG library used instead of NC. I run thousands of scripts without problems with the ipc layer. However, you should be aware of several items when running scripts from crontab so that you can avoid message queue id conflicts: 1) create a unique working directory for each invocation of a script I make a working dirctory such as /tmp/.workdir.$$ where "$$" is the csh syntax from the process id of the script, do my processing, and then remove the directory. This way, only one script executes per directory to avoid conflicts with the gemglb.nts. 2) Crons are not attached to a tty, so the message queue created is based on the process id of the parent script and not the tty. Uou have to ensure that gpend is run, especially if an error occurs in the script since no other process will be able to use that messagequeue and you could eventually run out of message queues. As you note, the _vg, _gf, _nc versions avoid the message queues, so that element of administration is easier- though overlays are not as easy as you have found. Steve Chiswell Unidata User Support > > Unidata GEMPAK Support wrote: > > >Mark, > > > >Yes, there is a utility to concatinate the vgf files. > > > > > > > >>vgappend > >> > >> > > > > Usage: vgappend file1 file2 [...fileN] > > Where file1 is the existing output file, > > and file2...fileN are the existing input files. > > > >I concatinate the mesocyclone products from individual sites such as: > >cd $VGDIR > >set VGFILES=`/opt/bin/find [A-Z]*.vgf -cmin -60 -print` > >if ( $#VGFILES > 0 ) then > > if ( -e .allmeso.vgf ) rm .allmeso.vgf > > cp $VGFILES[1] .allmeso.vgf > > shift VGFILES > > if ( $#VGFILES > 0 ) vgappend .allmeso.vgf $VGFILES > > mv -f .allmeso.vgf allmeso.vgf > >endif > > > > > >The above copies the contents of the first file in the list to a new file, > >and then concatinates the rest of the data in vgappend as an example > >since you may not want to clobber your original vgf file. > > > >Steve Chiswell > >Unidata User Support > > > > > > > >>Hi Chiz, > >> > >>Thanks for your reply: I migrated to GEMPAK5.10.3 and it provides a > >>number of features I need. > >> > >>I have an on-going problem where the communication between gpmap and > >>gplt locks-up. It could be the IPC message system has reached a limit > >>since a number of other processes use the service (same user name so I > >>can't disentangle easily). > >> > >>So, I compiled gpmap_ps. The problem is, I want to process 6 vgf > >>files. Is there any way? > >>It doesn't look like you can enter: > >> > >>VGFILE = file1.vgf;file2.vgf > >> > >>Is there a utility to concatenate the vgf files so that I can enter > >>something like: > >> > >>VGFILE = group_of_files.vgf > >> > >>Thanks, > >> > >>Mark Hughes > >>Global Weather Dynamics, Inc. > >> > >>Unidata GEMPAK Support wrote: > >> > >> > >> > >>>Mark, > >>> > >>>The SIGWX tools are done by NCEP in support of requirements of the AWC. > >>> > >>>You can browse through the change documents at: > >>> > >>>http://www.nco.ncep.noaa.gov/sib/nawips/ > >>> > >>>The 5.9.4 release notes mention a major restructuring of placement of > >>>features as defined in the $GEMTBL/pgen/settings.tbl and placement.tbl > >>> > >>>Since these are operational tools, I don't have much background in the > >>>formatting of > >>>display elements and requirements that have lead to those changes. Perhaps > >>>AWC would have > >>>someone that can inform you of that product. > >>> > >>>Steve Chiswell > >>>Unidata User Support > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>>>Institution: Global Weather Dynamics, Inc > >>>>Package Version: GEMPAK5.8.3a > >>>>Operating System: Linux Intel > >>>>Hardware Information: HP Pavillion > >>>>Inquiry: Hi, > >>>> > >>>>I am using the SIGWX package to display BUFR data. > >>>>My customer has a preference for the older style display of CAT > >>>>information in a box rather than as a label on the map itself. > >>>> > >>>>I see that at one time this was the standard approach. I would like to > >>>>investigate adapting the older routine for use in the version I am using > >>>>(5.8.3a). I couldn\'t find release notes which would tell me when this > >>>>occurred. Could you tell me what is the last release which used CAT > >>>>boxes? > >>>> > >>>>Thanks, > >>>> > >>>>Mark Hughes > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>>Ticket Details > >>>=================== > >>>Ticket ID: BPO-815416 > >>>Department: Support GEMPAK > >>>Priority: Normal > >>>Status: Closed > >>> > >>> > >>> > >>> > >>> > >> > >> > > > > > >Ticket Details > >=================== > >Ticket ID: BPO-815416 > >Department: Support GEMPAK > >Priority: Normal > >Status: Closed > > > > > > > > Ticket Details =================== Ticket ID: BPO-815416 Department: Support GEMPAK Priority: Normal Status: Closed
Attachment:
Makefile
Description: Binary data