IN2P3-Forge: Issueshttps://forge.in2p3.fr/https://forge.in2p3.fr/favicon.ico?16780521162011-02-15T12:02:22ZIN2P3-Forge
Redmine Pipelet - Task #1336 (New): Add comments in multiplexhttps://forge.in2p3.fr/issues/13362011-02-15T12:02:22ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>The code of multiplex is hard to read/lacks for comments. Variable Names, data structure may be rethought.</p> APCScheduler - Feature #1299 (New): Utilisation de python, numpy, scipy, ...https://forge.in2p3.fr/issues/12992011-02-10T11:06:35ZColley Jean-Marcjcolley@lpnhe.in2p3.fr
<p>Les librairies scientifiques (numpy, scipy, matplotlib) de python ne sont pas disponibles sur les workers. de la grille. APCScheduler doit offrir la possibilité d'utilisr les scripts python ayant recours à ces librairies.</p> LC2 - Feature #1255 (Assigned): Provide proper Shared-Object name to librarieshttps://forge.in2p3.fr/issues/12552011-01-17T13:39:12ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-10-09T11:52:11+0200<br />rpmlint indicates invalid-soname for libraries<br /><pre>
E: invalid-soname /opt/daq-MCH/LC2/lib/libLC2common.so libLC2common.so
E: invalid-soname /opt/daq-MCH/LC2/lib/libLC2version.so libLC2version.so
E: invalid-soname /opt/daq-MCH/LC2/lib/libLC2event.so libLC2event.so
E: invalid-soname /opt/daq-MCH/LC2/lib/libLC2config.so libLC2config.so
</pre><br />No version number is taken into account.</p>
<p>Adding a version number in the <br /><code>daq-MCH-LC2-1.3.2/LC2/config/sys/Linux-g++.gmk</code><br />Makefile is not enough.</p>
<p>the symbolic link to unversionned library must be provided as well. <br />See<br />[[<a class="external" href="http://www.gnu.org/software/libtool/manual/libtool.html">http://www.gnu.org/software/libtool/manual/libtool.html</a>]]</p> LC2 - Feature #1254 (Assigned): Add Tooltip for the three LC2 modeshttps://forge.in2p3.fr/issues/12542011-01-17T13:34:50ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-10-04T21:44:43+0200<br />The meaning of the three LC2 modes being unclear, <br />a concise explanation is wanted. <br />It could go along the following lines:<br />Edit : prepare Crocus command file for storage <br />Emulate : prepare Crocus command file, simulate sending it and load the corresponding result file for postmortem analysis <br />Send : prepare and send Crocus command files, ask Crocus for result file and display it in the Load cmd/result tab</p> LC2 - Bug #1253 (Assigned): use / harden checks of getlinehttps://forge.in2p3.fr/issues/12532011-01-17T13:32:58ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-09-15T12:32:38+0200<br />we have 4 cases of<br /><pre>
warning: unused variable ‘isReadingCorrect’
</pre></p>
<p>that could be further exploited.</p>
<pre>
src/CrocusConfig.cxx: In ScanFromResult?(std::istream&, Crocus::ui32&)
src/CrocusConfig.cxx:280:8:
src/CrocusConfig.cxx: In ScanFromFeC2 (const std::string&)
src/CrocusConfig.cxx:648:12:
src/CrocusConfig.cxx: In ReadPedestalConfig? (const std::string&, double)
src/CrocusConfig.cxx:840:10:
src/CrocusConfig.cxx:848:10:
int llPatchBus1 = -1;
in.get (first);
while (first == '/') {
bool isReadingCorrect = getline (in, getstr);
in.get (first);
}
in.unget();
while (!in.eof ()) {
in >> Bus_Patch_Abs >> Manu_Id >> Channel >> Mean >> Sigma;
if (in.eof ()) break;
bool isReadingCorrect = getline (in, getstr);
</pre><br />Further case are not exploited:<br /><pre>
source/config/src/utilities.cxx: isReadingCorrect = getline (stream, line);
source/config/src/utilities.cxx: isReadingCorrect = getline (stream, line);
source/gui1/src/ListView.cxx://[VLA]// in.getline (line, 120);
source/version/src/CrocusVerHeaderFile.cxx: in.getline (line, 120);
</pre>
<p>... and should be !</p> LC2 - Bug #1252 (Assigned): What is the proper Parsing Algorithm?https://forge.in2p3.fr/issues/12522011-01-17T11:34:39ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-09-15T01:02:01+0200In<br />bool FrtCrtConfig::ScanFromResult(std::istream& stream, ui32& key)</p>
<p>the following code block lead to rejection of a seemingly well formed result file:<br /><pre>
// If event padder, read dataKey
if ( payLoadLength % 2 == 1 ) {
result = ReadNextValue(dataKey, stream);
if ( ! result ) return false;
}
</pre></p>
<p>payLoadLength was 75=0000000x4b<br />instead of 0x00000032 for a good case</p> LC2 - Bug #1235 (Assigned): shared-lib-calls-exithttps://forge.in2p3.fr/issues/12352011-01-12T00:47:28ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-09-13T04:23:04+0200<br /><pre>
/opt/daq-MCH/LC2/lib/libLC2config.so
/opt/daq-MCH/LC2/lib/libLC2version.so
</pre></p>
<p>These library packages calls exit() or _exit(), probably in a non-fork() context. Doing so from a library is strongly discouraged - when a library function calls exit(), it prevents the calling program from handling the error, reporting it to the user, closing files properly, and cleaning up any state that the program has. It is preferred for the library to return an actual error code and let the calling program decide how to handle the situation.</p>
<p>The concerned classes are:<br /><pre>
LC2/source/version/src/CError.cxx
LC2/source/config/src/FrtCrtConfig.cxx
LC2/source/config/src/utilities.cxx
</pre></p> LC2 - Feature #1232 (Assigned): Include parseCrocusVersion in the Crocus Threadhttps://forge.in2p3.fr/issues/12322011-01-12T00:18:42ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-08-25T23:34:45+0200</p> LC2 - Feature #1231 (Assigned): Include boot in the Crocus Threadhttps://forge.in2p3.fr/issues/12312011-01-12T00:16:44ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-08-25T23:33:54+0200</p> LC2 - Feature #1156 (Assigned): query Marc Status generation & validation optionshttps://forge.in2p3.fr/issues/11562010-12-07T16:37:58ZLafage Vincentlafage@ipno.in2p3.fr
<p>2010-07-15T19:32:22+0200<br />modify & test<br /><pre>crocus_generator</pre> & <pre>crocus_file_validator</pre></p> Oval - Bug #1083 (New): oval log -guihttps://forge.in2p3.fr/issues/10832010-11-16T11:42:19ZChamont David
<p>The adaptative nature of the command line interface, which choose dynamically the level of details of its output depending on the number of targets, is generally a source of confusion for the graphic user interface, which relies on those outputs and would prefer a single fixed output format for each command.</p>
<p>For example, for the "oval log" command, should we consider a "-gui" option which switch off the dynamic output and produce something specific for the gui ? Should we implemetn such an option for any command ? Or shoudl we rather focus on giving an API for the gui, so that it does rely on the cli (Command Line Interface) (and incidentally executes a lot more faster...).</p> Oval - Bug #1082 (New): NO DIFFhttps://forge.in2p3.fr/issues/10822010-11-16T11:38:30ZChamont David
<p>When "oval log" displays "NO DIFF", it is not easy for the newcomer to distinguish between "there is no difference" or "the diff command has not been applied". We should investigate how to make it more clear.</p> SphereLib - Bug #1003 (New): ps detectionhttps://forge.in2p3.fr/issues/10032010-10-15T13:51:07ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>Perform ps detection in 2 steps if sigma < 20</p> SphereLib - Feature #959 (New): intégrer ginphttps://forge.in2p3.fr/issues/9592010-10-06T11:30:50ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>interfacer ginp depuis python</p> SphereLib - Bug #938 (New): Construction icosahedronhttps://forge.in2p3.fr/issues/9382010-09-29T11:32:48ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>Utiliser le module fortran pour la compilation de icosahedron.f</p> Oval - Feature #923 (New): oval resethttps://forge.in2p3.fr/issues/9232010-09-20T17:16:27ZChamont David
<p>Pour remettre à zéro les log files, éventuellement effacer les locks de vieux runs mal interrompus, etc.</p> Oval - Bug #905 (New): Use of parenthesis in diffline regular expressionhttps://forge.in2p3.fr/issues/9052010-09-14T18:17:51ZChamont David
<p>In case one is using litteral parenthesis in a diffline regular expression, backquoted or between square brackets, FilesParser wrongly think that the user has included parenthesis in the expression so to delimit the region of interest, and consequently it does not add default parenthesis so to select the whole line as region of interest. As a consequence, the selection of this "region of interest" always return "1" whatever the real content.</p>
<p>For the user, a fast and dirty way to avoid the problem is to add himself the parenthesis.</p> SphereLib - Feature #890 (New): spline with order != 3https://forge.in2p3.fr/issues/8902010-09-06T15:08:41ZLe Jeune Maudelejeune@apc.in2p3.frSphereLib - Task #767 (New): Mise au propre des include de la lib https://forge.in2p3.fr/issues/7672010-06-16T11:36:41ZLe Jeune Maudelejeune@apc.in2p3.frSphereLib - Feature #755 (New): Parallélisation de cat2maskhttps://forge.in2p3.fr/issues/7552010-06-14T16:34:59ZLe Jeune Maudelejeune@apc.in2p3.frSphereLib - Feature #754 (New): Parallélisation de la routine apodize_maskhttps://forge.in2p3.fr/issues/7542010-06-14T16:33:47ZLe Jeune Maudelejeune@apc.in2p3.frSphereLib - Task #742 (New): Import des binaires fitshttps://forge.in2p3.fr/issues/7422010-06-09T19:24:03ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>Importer les binaires C++/fits. Harmoniser les paramètres avec ceux des routines python.</p> Pipelet - Feature #734 (Assigned): Rapport Latexhttps://forge.in2p3.fr/issues/7342010-06-07T19:06:18ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>Intégrer le code de génération automatique de rapport latex</p> Pipelet - Bug #723 (Assigned): Timers on lockhttps://forge.in2p3.fr/issues/7232010-06-07T18:55:32ZLe Jeune Maudelejeune@apc.in2p3.fr
<p>Timers on Lock are definitely needed</p> ENX - Support #660 (New): Documentationhttps://forge.in2p3.fr/issues/6602010-05-27T10:55:30ZLegay EricEric.Legay@csnsm.in2p3.fr
<p>Déplacer toute la doc vers la Forge !!!!</p> ftsmonitor - Feature #631 (New): History of a DN accesshttps://forge.in2p3.fr/issues/6312010-05-19T15:42:41ZSchwarz Lionel
<p>This could be used to trace access by DN in case of security issue</p> Oval - Feature #597 (New): YAML comme langage de configuration ?https://forge.in2p3.fr/issues/5972010-05-11T13:36:16ZChamont David
<p>Réfléchir à l'opportunité d'utiliser YAML comme langage de configuration.</p> Smurf - Feature #202 (New): Smurf::DB::Getopt add fping to smelecthttps://forge.in2p3.fr/issues/2022009-10-21T16:46:45ZWernli Fabienwernli@in2p3.fr
<p>It should be possible to add -ping flag to e.g. smelect<br />to output only servers which are alive</p> RlsngCons - Feature #187 (New): Abort limit/delete/ackhttps://forge.in2p3.fr/issues/1872009-10-07T13:08:32ZWernli Fabienwernli@in2p3.fr
<p>Ctrl-G should abort l/A/D edition</p> RlsngCons - Feature #186 (New): Flood limithttps://forge.in2p3.fr/issues/1862009-10-07T13:07:05ZWernli Fabienwernli@in2p3.fr
<p>Limit flood for pasted test, e.g. like Ctrl-K in irssi</p> Oval - Feature #174 (New): oval run -ohttps://forge.in2p3.fr/issues/1742009-10-02T14:47:08ZChamont David
<p>We should have a way to generate a script, able to run a given target in the conditions defined in OvalFile, without any more need for oval. In such a case, of course, the output could not be inspected and compared to a reference. This would make Oval usable as a runtime scripts generator, and help use Oval in a group where it is not an official tool.</p> Oval - Bug #172 (New): Automatic variableshttps://forge.in2p3.fr/issues/1722009-10-02T14:24:53ZChamont David
In several use-cases, including interaction with batch/grid tools, it would help if Oval was setting automatically few variables. A first set could be :
<ul>
<li>OVAL_CWD : the unix directory where the oval command has been invoked.</li>
<li>OVAL_ENV : name of the environment where the current target is staying.</li>
</ul> RlsngCons - Feature #93 (New): Touche pour forcer le renouvellement de cachehttps://forge.in2p3.fr/issues/932009-03-31T10:16:29ZWernli Fabienwernli@in2p3.fr
<p>Par exemple $ comme pour mutt <img src="/plugin_assets/redmine_wiki_extensions/images/wink.png" alt=";)"></p> RlsngCons - Feature #87 (New): Limiter la taille du cachehttps://forge.in2p3.fr/issues/872009-03-30T15:32:25ZPuel Mattieumpuel@cc.in2p3.fr
<p>En plus des filtres de cache, il faut pouvoir définir une taille de fenêtre maximale du cache à partir de laquelle il commence à supprimer les anciens messages au fur et à mesure de l'arrivée de nouveaux.</p> RlsngCons - Feature #86 (New): Tris dynamiqueshttps://forge.in2p3.fr/issues/862009-03-30T15:31:08ZPuel Mattieumpuel@cc.in2p3.fr
<p>On doit pouvoir trier dynamiquement les messages en fonction d'un champs.</p> Oval - Feature #37 (New): Automatically select the shell flavorhttps://forge.in2p3.fr/issues/372009-02-16T19:44:59ZChamont David
<p>Oval should recognize the current user shell, and make it the default one for oval commands.</p> Oval - Feature #36 (New): Extend tasks with begin/end directory commandshttps://forge.in2p3.fr/issues/362009-02-16T19:42:43ZChamont David
<p>Perhaps, the administrators should be given the possibility to define a command to be executed each time Oval enter a new directory, and another one each time Oval exit from a directory.</p> Oval - Feature #35 (New): Give less importance to the build stephttps://forge.in2p3.fr/issues/352009-02-16T19:41:03ZChamont David
<p>The documentation always mention the Oval buidl step as the usual way of working. This is not necessarily true, and could make the doc not so easy to read, because the build step is not the simpler, especially for the configuration.</p>
<p>The documentation should be reviewed so that the build step is described separately and only at the second stage.</p>
<p>We should even reconsider how important it is to have the build step in Oval.</p> Oval - Feature #34 (New): Extend input/output controlhttps://forge.in2p3.fr/issues/342009-02-16T19:27:17ZChamont David
<p>There is discrepancy in the way input and output are considered: input must be a file, and output is a channel. There should be way to make the diff on some output file, and there should be a way to send input to a program via a channel.</p> Oval - Feature #33 (New): Support for gzipped ref fileshttps://forge.in2p3.fr/issues/332009-02-16T19:23:37ZChamont David
<p>By André Holzner :</p>
<p>Hello David,</p>
<p>I produced an OVAL log file which is 1.3 MBytes. I don't<br />dare to commit it to CVS. If I gzip it (with -9), it<br />is slightly larger than 27'000 bytes, using bzip2<br />it even shrinks to less than 20'000 bytes.</p>
<p>Thus my question: Would it be possible to add support<br />for gzipped or bzipped ref files ? I think I could achieve<br />the same result using a combination of a script which runs<br />the actual executable and compresses the output and a special diff command... but I guess it should not be<br />too difficult to add a bzcat / zcat somewhere in oval<br />where the diff is performed...</p>
<p>What do you think about it ?</p>
<p>best regards,</p>
<p>Andre</p> Oval - Bug #32 (New): Write a more detailed configuration dochttps://forge.in2p3.fr/issues/322009-02-16T19:18:24ZChamont David
<p>Especially the description of tools.</p> Oval - Feature #31 (New): Apply a command to a subdirectory subset ?https://forge.in2p3.fr/issues/312009-02-16T19:14:14ZChamont David
<p>Do we want to enable something such as "oval prod SubDir1/Test2" ?</p> Oval - Feature #30 (New): Use environments as targets to the step commandshttps://forge.in2p3.fr/issues/302009-02-16T19:12:58ZChamont David
<p>Would be an additionnal way to execute a subset of the targets. Or do we need a "set" concept ?</p> Oval - Feature #29 (New): The build tool should handle directories ?https://forge.in2p3.fr/issues/292009-02-16T19:06:25ZChamont David
<p>Scram is able to receive a directory as argument. Should we modify Oval so that it takes such a possibility into account ?</p> Oval - Bug #28 (New): "oval validate" should remove differences in "oval log"https://forge.in2p3.fr/issues/282009-02-16T19:01:54ZChamont David
<p>As said in the title...</p> Oval - Bug #27 (New): Some commands should not require a current valid directoryhttps://forge.in2p3.fr/issues/272009-02-16T18:51:50ZChamont David
<p>For example, in a scram configuration, one should be able to type "oval version" and have a valid output, even if he is not within a valid scram project.</p> Oval - Bug #26 (New): Commands options not well propagated, when declared within an OvalFilehttps://forge.in2p3.fr/issues/262009-02-16T18:50:15ZChamont David
<p>Intro: so to ensure user interface and commands consistency, each single feature is implemented in a single command, and when a command needs another command feature, it is internally instanciating this command and use it. Behind this nice concept, since the internal command is not instanciated in the same way as the main command, we must propagate to it some "fake command line" options, and fakely calls its functions, as would do the main oval framework with the top level command. This fake calls are not done automatically, but case by case and "by hand" by the top command. This is why there could be and there is some<br />errors or awkward situation.</p>
<p>The general bug: in the current implementation, when a commmand option is set within an OvalFile, it is not<br />correctly propagated to the internals commands.</p>
<p>Example: if I set "-sh" for the runtime command, it will not be propagated to "environnement" command, and the<br />resulting output will be a mix of csh and sh commands.</p>
<p>Solution: review the use of internal commands, and the way they are managed, so that they are in the same state if they are instanciated as user commands or as internal commands.</p> Oval - Bug #25 (New): Temporary files after a crashhttps://forge.in2p3.fr/issues/252009-02-16T18:46:14ZChamont David
<p>If an oval process is killed or crashes, we should check if the temporary files remain in the current directory or not, which may confuse subsequent oval runs.</p> Oval - Bug #22 (New): Deliver messages when ISA failshttps://forge.in2p3.fr/issues/222009-02-16T18:22:12ZChamont David
<p>When there is a perl failure, for example when ISA fails because perl does not find the relevant base package, it would be nice if oval could catch the event and deliver the defered messages before exiting.</p> Oval - Bug #21 (New): What happens when a subdirectory is reused in several environments ?https://forge.in2p3.fr/issues/212009-02-16T18:16:49ZChamont David
<p>Undefined behavior...</p>