Hi, Chad Matsalla wrote: > > Greetings! > > I sent this to the basedb-devel mailing list last week but it didn't > turn up so I'm trying again. > > > I'm using BASE-2.4.0 and tomcat 6.0.14. I also tried base-2.4.0 and > tomcat 5.5.20. > > I'm trying, as usual, to run BASE1 plugins. I was close in BASE-2.3.2 in > that the only problem seemed to be that the last line of the bioassay > set wasn't written into the stdin.txt file. > > In BASE-2.4.0, I have several problems.
Don't give up, we have reports from users who run BASE1 plug-ins on a BASE2 server. > 1. It seems that a BASE1 plugin runs before the data was successfully > generated out of the database. (see explanation below) Hm, this is strange. Can you try to use the resolution to item 2 below and examine the file in the temporary work directory? The problem of saving stdin.txt is reported as a bug and will be resolved in 2.4.1 (cf. http://base.thep.lu.se/ticket/739) > 2. The basefile is generated into a tomcat temp directory and is > unlinked before I can analyze it. The temporary directory is automatically deleted after a successful end of the plug-in. Successful in this context means that the plug-in either finalizes its tasks or makes an graceful exit if an exception is encountered. If you remove the deleteDir() call in the plugin (line 682, http://base.thep.lu.se/browser/trunk/src/plugins/core/net/sf/basedb/plugins/Base1PluginExecuter.java#L682) all files generated during plugin execution will remain in tomcats temp directory. > 3. I can't empty the trash Sorry, no resolution on this item from me. Someone else might have a clue. > Plugin runs before data is generated > ------------------------------------ > I call a perl executable on the server - this executable reads the > stdout from the base server and then calls R. Because of problem (2) > (described above), I've modified the perl program to also write the > output to /tmp. > > In /tmp I see a file containing: > > BASEfile > fname rnormv1.R > norm.method p > do.bgcorr 0 > ngrid.r.in 12.0 > ngrid.c.in 4.0 > nspot.r.in 18.0 > (and a whole bunch more stuff that is correct but the last two lines I > see are:) > 863 18058 A006151_01 565 398 0 444 282 243 146 0 197 > 864 85459 At30009369 620 441 0 397 276 248 1 > > It seems that I'm only seeing the first 864 rows of reporter data and > the last line is obviously truncated. > > When I export data from the Bioassay set in the Experiments viewer I get > a proper basefile with 31200 rows of reporter data. > > > File is deleted before I can Examine It > ---------------------------------------- > It appears that the stdin.txt file containing the basefile intended for > the Base1PluginExecutor is being written to a tomcat tempfile somewhere. > I looked all through base.userfiles with commands like: > > userfiles$ find . -type f -mtime -1 -exec ls -l {} \; > > I found the r-debug.txt that the R-program generated (indicating that > rows were missing from the basefile) but no XXXXXXXstdin.txt (where > XXXXXXX is the BASE serial number). > > This behavior changed since 2.3.2 and I can't find any information > regarding this change in TRAC. In previous versions checking the 'Leave > stdin.txt' in the plugin Configuration, well, resulted in stdin.txt being > left behind. > > > Emptying the trash > ------------------ > In my trashcan I have 1.6 gazillion files. I followed the instructions > here: > http://base.thep.lu.se/chrome/site/doc/html/userdoc/trashcan/trashcan.viewdependencies.html > > > But whenever I try to delete items using the 'Empty Trash...' command I > get: "0 of xxx items removed. xxx items are used and can't be removed.' > xxx is some number. > > This is interesting because if I filter by Item Type 'File', click on > the little icon on the left (looks like a paper document with a piece of > a chain on it) see no indication that it has dependencies. > > If I select a single File item from the Trashcan and click 'Delete > Permanently' I get the "can't be removed" message. > > If, in the Properties page for the same file I click 'Delete > Permanently....' no error message appears and when I return to the > 'Trashcan' page the item is gone and the total count of hits is > decremented. > > In addition, the command: > userfiles $ find . -type f | wc > returns one less. This indicates that the delete was successful. > > Do I have to delete files one at a time? > > > Summary > ------- > > Can somebody please help me understand the best way to: > > 1. Restore the behavior that I saw in previous versions of BASE2 where > the stdin.txt that was intended to be used for a base1 plugin appear on > the filesystem? I can then use it to troublshoot the R program that does > my work. > > 2. Run base1 plugins? I can't seem to get the entire basefile out of > base. > > 3. Empty my trash. > > > Thank you for any help you can provide. > > Chad Matsalla > Plant Biotechnology Institute, National Research Council of Canada > Saskatoon, Saskatchewan -- Jari Hakkinen, PhD Complex Systems Division mailto:[EMAIL PROTECTED] Department of Theoretical Physics phone: +46 (0)46 2229347 Lund University fax: +46 (0)46 2229686 Solvegatan 14a, SE-223 62 Lund, Sweden http://www.thep.lu.se ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ basedb-devel mailing list basedb-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/basedb-devel
