Lars - Thanks for all of your changes. Mohan and I will download your branch and start testing your simulator in the next day or so.
If all goes well, hopefully by next week, you can merge your new simulator into the main OpenHPI branch. The goal is to do a code freeze on the main branch on 6/14/2010 (Monday) - in anticipation of the 2.15.0 release. Of course, that date can slip somewhat if we run into issues - but I just wanted to make you aware of it. BTW - did you test the building of RPMs since you made the name change? Uli, will you be able to do any more testing of Lars' simulator? --michael Regards, Michael Bishop Enterprise Servers and Storage (ISB Linux/Telco) Hewlett-Packard Company 3404 E. Harmony Rd. Bldg. 5L, Post B7, Mailstop 42 Fort Collins, CO 80528-9599 Phone: 970-898-4393 E-Mail: [email protected] > -----Original Message----- > From: Lars Wetzel [mailto:[email protected]] > Sent: Thursday, June 03, 2010 2:52 PM > To: [email protected] > Cc: Bishop, Michael (ISB Linux/Telco); Devarajulu, Mohanasundaram; > Kleber, Ulrich (NSN - DE/Munich) > Subject: Re: [Openhpi-devel] New simulator testing > > > Hi Michael, Mohan and Uli, > > ok following changes were made: > > 1. and 3. debug and logfile > The logfile generation is now deactivted by default. > > 4. ep of announcements > Fix of simulation.data.example > > 6. naming > Change the naming to dynamic_simulator (library libdyn_simulator) > It works on my machine, but due to some problems with svn (I > had to commit in > 3 steps) I hope it works also with the data inside the repository. > Furthermore it is and was the most critical part of the changes. > > 2. new <-> old > it's still your part ;-) > > 5. timestamp of announcement > fault or not fault, it has to be discussed if it is really > necessary to work > with original timestamp in the announcements. > > soa_plugin: > Testcases > Sensor/saHpiSensorReadingGet/GetAndVerify > Sensor/saHpiSensorThresholdsGet/GetThresholds > Will now pass. > The reason was a inconsistence of the data. I enhanced the > hpigensimdata > client to overcome it. (DataFormat.IsSupported was set on > SAHPI_TRUE but > Reading.IsSupported returns SAHPI_FALSE) > I don't understand how the test cases can pass in your > system. Maybe it was a > temporarily bug? But nevertheless the client will throw a > warning if it > happens again. > > Sensor/saHpiSensorEventMasksSet/SensorEnableChangeEvent > Wasn't supported by old and new simulator. > Now it is supported in the new simulator and had a small bug. - Fixed > > Inventory/saHpiIdrAreaHeaderGet/TraversalByAreaType > Was a bug -> fixed > > Inventory/saHpiIdrFieldSet/SetField > {COOLING_DEVICE,x} and FANs > Fails due to wrong data -> Field has wrong Type entry > Can you please verify it with original hardware > > Inventory/saHpiIdrAreaAddById/AddAreaById > SAHPI_FIRST_ENTRY is a valid AreaId and can be used with this > function, > provided the IDR does not have a pre-existing Area with that AreaId. > In the data the areas starts with the Id 0 == > SAHPI_FIRST_ENTRY. So I think > the issue is more a problem of the test case?! > > HotSwap/saHpiAutoExtractTimeoutGet/GetTimeout > Was a Typo, is now fixed > > atca_data: > Sorry is still open: > POWER_CYCLE -> Event generation is missing > Textbuffer doesn't support binary data > > Hope I didn't bring in some new faults. > > Best Regards > Lars > > On Tuesday, 1. June 2010 17:17, Bishop, Michael (ISB > Linux/Telco) wrote: > > Lars - > > > > I'm fine with your plan for the log files. > > > > Assuming you fix the issues pointed out by Mohan, I'm also > fine with making > > the new simulator the default. Anyone have objections to this? > > > > I'm don't think the timestamp issue has to be fixed for the > first release. > > > > For a name - I still like dynamic_simulator (or > dyn_simulator for short > > name.) > > > > Please let us know when you have posted a new version for > us to test. I'd > > like to complete the testing this week - if possible, and > then merge the > > new plugin to the top-of-trunk next week, prior to starting > our 2.15.0 > > release cycle. > > > > --michael > > > > Regards, > > Michael Bishop > > Enterprise Servers and Storage (ISB Linux/Telco) > > Hewlett-Packard Company > > 3404 E. Harmony Rd. Bldg. 5L, Post B7, Mailstop 42 > > Fort Collins, CO 80528-9599 > > Phone: 970-898-4393 > > E-Mail: [email protected] > > > > > -----Original Message----- > > > From: Lars Wetzel [mailto:[email protected]] > > > Sent: Monday, May 31, 2010 9:48 AM > > > To: [email protected] > > > Cc: Kleber, Ulrich (NSN - DE/Munich); Devarajulu, Mohanasundaram; > > > Bishop, Michael (ISB Linux/Telco) > > > Subject: Re: [Openhpi-devel] New simulator testing > > > > > > > > > Hi, > > > > > > Thank you for testing! > > > > > > I plan to change the following points - please comment it, if > > > you think it is > > > the wrong direction. > > > > > > 1 and 3.) Both pointed to one issue - the new logfiles. > > > Background - it is the same mechanism as in the ipmidirect > > > plugin. I plan to > > > deactivate the generation of the log files by default (as it > > > is done in the > > > ipmidirect plugin). The information in the files is only > > > needed for debugging > > > and since you test the version from the development branch it > > > was activated > > > by default. > > > Errors are written to /var/log/messages - as at the other > > > plugins. This means > > > there will be no restrictions, if I deactivate the additional > > > logfiles. > > > > > > I don't see the necessity to spend development effort for > > > seperating logfiles > > > if two simulator plugins are started. First of all the > option will be > > > deactiveted, see above. Furthermore we are talking about a > > > simulator. This > > > means it runs temporarily to test some data or clients. I > > > don't see the use > > > case to have two seperate logfiles for each simulator > > > instance. You can > > > easily re-start the daemon with one plugin, if you want to > > > debug an issue. > > > > > > 4.) Entity path: It looks to me like a bug, I will change the > > > data and re-test > > > it. > > > > > > 2.) old and new simulator by default - it is up to you. > > > But the simulation data file is the only additional > > > configuration task and > > > should work by default. > > > > > > 5.) Timestamps: It was easier to implement one function for > > > the annunciator > > > and to spare the parsing of the timestamps. At the moment the > > > "normal" > > > annunciator functions are used at import, which means the > > > annunciator writes > > > it's own timestamp. > > > > > > 6.) Name: > > > I can change the name of the plugin, but I will not change > > > the names of the > > > classes (NewSimulatorXXX) and files (new_sim_xxx). For this > > > it is too late. > > > > > > Proposal: DataSimulator? > > > > > > Best regards > > > Lars > > > > > > > > > On Thursday, 27. May 2010 11:00, Kleber, Ulrich (NSN - > > > > > > DE/Munich) wrote: > > > > Hi, > > > > here is my two cents on some of these issues: > > > > > > > > 1. I agree that default for debug should be off > > > > > > > > 2. I think the oldsimulator should stay available. > > > > The new simulator needs some configuration to do, which > > > > could be complex for a newcomer. > > > > But I think the newsimulator could have a fancy name, which > > > > expresses the major benefits it provides, e.g. > dynamic-simulator > > > > or configurable-simulator. > > > > Maybe we should rename the old simulator to simple-simulator? > > > > > > > > 3. I would prefer the possibility to have separate > logfiles if you > > > > activate two plugins. > > > > > > > > Cheers, > > > > Uli > > > > > > > > > -----Original Message----- > > > > > From: ext Bishop, Michael (ISB Linux/Telco) > > > > > [mailto:[email protected]] > > > > > Sent: Wednesday, May 26, 2010 5:55 PM > > > > > To: '[email protected]'; Lars Wetzel > > > > > Cc: Devarajulu, Mohanasundaram > > > > > Subject: Re: [Openhpi-devel] New simulator testing > > > > > > > > > > Lars - > > > > > > > > > > I agree with Mohan that debug messages should not be turned > > > > > on by default. > > > > > > > > > > I'm not sure the old simulator should be removed from > > > > > openhpi.conf. Perhaps it could be commented out - and the > > > > > new simulator used as the default plug-in. Does anyone else > > > > > have an opinion about this? If the new simulator is going to > > > > > be the default plug-in - it needs to be rock solid. > > > > > > > > > > Please investigate Mohan's comment about the log file (#3 > > > > > below). We do not want the new simulator to be confusing to > > > > > OpenHPI users, especially if it becomes the default plug-in. > > > > > > > > > > Also - please investigate 4 and 5 and let us know of any > > > > > changes you decide to make. > > > > > > > > > > Thanks! > > > > > > > > > > --michael > > > > > > > > > > Regards, > > > > > Michael Bishop > > > > > Enterprise Servers and Storage (ISB Linux/Telco) > > > > > Hewlett-Packard Company > > > > > 3404 E. Harmony Rd. Bldg. 5L, Post B7, Mailstop 42 > > > > > Fort Collins, CO 80528-9599 > > > > > Phone: 970-898-4393 > > > > > E-Mail: [email protected] > > > > > > > > > > > -----Original Message----- > > > > > > From: mo...@fc [mailto:[email protected]] > > > > > > Sent: Tuesday, May 25, 2010 5:06 PM > > > > > > To: Lars Wetzel > > > > > > Cc: [email protected] > > > > > > Subject: [Openhpi-devel] New simulator testing > > > > > > > > > > > > > > > > > > Hi Lars, > > > > > > > > > > > > New simulator is having better results as for as > hpib-test is > > > > > > concerned. > > > > > > Here are some of my observations with the testing. > > > > > > > > > > > > 1. Looks like the debug option is turned on by default > > > > > > in the new > > > > > > > > > simulator plugin. This could be off. > > > > > > > > > > > > 2. Once the new_simulator code is merged the old simulator > > > > > > > > > > information > > > > > > > > > > > could be removed from the openhpi.conf > > > > > > > > > > > > 3. I do not understand the need for the log file in the > > > > > > > > > > openhpi.conf. > > > > > > > > > > > More over even if I define two logfiles for 2 new_simulator > > > > > > entries, it > > > > > > takes only one. > > > > > > > > > > > > 4. The default simulation.data needs to have complete > > > > > > > > > > entity path for > > > > > > > > > > > every entity. But there are few that do not have the base > > > > > > path now. They > > > > > > are mis-interpreted now, based on the path defined in the > > > > > > openhpi.conf. > > > > > > They are {SYSTEM_BOARD,1}, {DISK_DRIVE,2} in the > > > > > > ANNUNCIATOR/RDR_DETAIL. > > > > > > They get converted to {SYSTEM_BOARD,X}, {DISK_DRIVE,X} > > > > > > (X is 9 in my > > > > > > > > > test case) as {SYSTEM_CHASSIS,1} is prefixed to these. > > > > > > > > > > These could be > > > > > > > > > > > seen in the output of hpigensimdata > > > > > > > > > > > > 5. timestamps are changed in the hpigensimdata output too. > > > > > > > > > > > > I will be doing more testing. Will report the results to > > > > > > > > > > you as I find > > > > > > > > > > > out more. > > > > > > > > > > > > Thanks > > > > > > Mohan > > > > > > > > > > > > -- > > > > > > mo...@fc <[email protected]> > > > > > > > > > > > > > > > > > > > -------------------------------------------------------------- > > > > > > ---------------- > > > > > > > > > > > > _______________________________________________ > > > > > > Openhpi-devel mailing list > > > > > > [email protected] > > > > > > https://lists.sourceforge.net/lists/listinfo/openhpi-devel > > > > > > > > > > -------------------------------------------------------------- > > > > > ---------------- > > > > > > > > > > _______________________________________________ > > > > > Openhpi-devel mailing list > > > > > [email protected] > > > > > https://lists.sourceforge.net/lists/listinfo/openhpi-devel > > > > > > -------------------------------------------------------------- > > > ------------- > > > > > > >--- > > > > > > > > _______________________________________________ > > > > Openhpi-devel mailing list > > > > [email protected] > > > > https://lists.sourceforge.net/lists/listinfo/openhpi-devel > > > > > > -- > > > ------------------------------- > > > Dipl. Wi.ing. > > > Lars Wetzel > > > Uttinger Str. 13 > > > 86938 Schondorf a. Ammersee > > > > > > Tel.: 0179-2096845 > > > Mail: [email protected] > > > > > > USt-IdNr.: DE181396006 > > > > > -------------------------------------------------------------- > ------------- > >--- > > > > _______________________________________________ > > Openhpi-devel mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/openhpi-devel > > -- > ------------------------------- > Dipl. Wi.ing. > Lars Wetzel > Uttinger Str. 13 > 86938 Schondorf a. Ammersee > > Tel.: 0179-2096845 > Mail: [email protected] > > USt-IdNr.: DE181396006 > ------------------------------------------------------------------------------ ThinkGeek and WIRED's GeekDad team up for the Ultimate GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the lucky parental unit. See the prize list and enter to win: http://p.sf.net/sfu/thinkgeek-promo _______________________________________________ Openhpi-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/openhpi-devel
