From chemical.aniket at gmail.com Tue Aug 1 10:30:16 2006 From: chemical.aniket at gmail.com (Aniket Kulkarni) Date: Tue, 1 Aug 2006 20:00:16 +0530 Subject: [Disksim-users] simulated events Message-ID: What are simulated events & what are sequences of event? Please inform as early as possible. with regards, aniket -------------- next part -------------- An HTML attachment was scrubbed... URL: From bucy at ece.cmu.edu Tue Aug 1 14:10:20 2006 From: bucy at ece.cmu.edu (John Bucy) Date: Tue, 1 Aug 2006 11:10:20 -0700 Subject: [Disksim-users] simulated events In-Reply-To: References: Message-ID: <4fc1e0430608011110r11c69c57oc03adb1d66744140@mail.gmail.com> On 8/1/06, Aniket Kulkarni wrote: > What are simulated events & what are sequences of event? > Please inform as early as possible. I'm not sure I understand what you're asking about... what is the context of this? john From hayashemtov at yahoo.com Wed Aug 9 05:26:14 2006 From: hayashemtov at yahoo.com (haya shem tov) Date: Wed, 9 Aug 2006 02:26:14 -0700 (PDT) Subject: [Disksim-users] some questions regarding Disksim Message-ID: <20060809092614.69084.qmail@web42409.mail.scd.yahoo.com> Dear Disksim users, I am new in disksim and I am trying to integrate disksim in a larger simulator Using disksim_interface.c. I am having some difficulties since I do not fully understand the API. Does anyone has more documentation on disksim interface functions, or more example programs? In detail, in the file syssim_driver.c when the function syssim_report_completion is called , how can I know which request was completed? Is there no ID filed in the Request structure? Best, Haya Shemtov --------------------------------- Want to be your own boss? Learn how on Yahoo! Small Business. -------------- next part -------------- An HTML attachment was scrubbed... URL: From sgovinda at cse.psu.edu Wed Aug 9 12:25:11 2006 From: sgovinda at cse.psu.edu (Sriram Govindan) Date: Wed, 9 Aug 2006 12:25:11 -0400 (EDT) Subject: [Disksim-users] some questions regarding Disksim In-Reply-To: <20060809092614.69084.qmail@web42409.mail.scd.yahoo.com> References: <20060809092614.69084.qmail@web42409.mail.scd.yahoo.com> Message-ID: hi shemtov, Request submission and completion at the IOdriver level is done in handle_new_request and iodriver_access_complete functions respectively in iodriver_disksim.c. From here you can track the complete flow. Hope this helps, sriram.G On Wed, 9 Aug 2006, haya shem tov wrote: > Dear Disksim users, > I am new in disksim and I am trying to integrate disksim in a larger simulator > Using disksim_interface.c. > I am having some difficulties since I do not fully understand the API. > Does anyone has more documentation on disksim interface functions, or more example programs? In detail, in the file syssim_driver.c when the function syssim_report_completion is called , how can I know which request was completed? Is there no ID filed in the Request structure? > Best, > Haya Shemtov > > --------------------------------- > Want to be your own boss? Learn how on Yahoo! Small Business. From xyao at cse.unl.edu Wed Aug 9 15:27:43 2006 From: xyao at cse.unl.edu (Xiaoyu Yao) Date: Wed, 9 Aug 2006 13:27:43 -0600 Subject: [Disksim-users] some questions regarding Disksim In-Reply-To: Message-ID: <200608091827.k79IRq5G003391@cse.unl.edu> Hi In syssim_driver.c, the next request is assumed to be dispatched until the last one has finished and reported by disksim. So here we don't need to know the request id when the syssim_report_completion() is triggered. In disksim 3.0, there is a syssim_driver2.c which shows a more useful example about using disksim as slave. The code is much longer than syssim_driver.c. But you can tailor it to your needs. Xiaoyu -----Original Message----- From: disksim-users-bounces at ece.cmu.edu [mailto:disksim-users-bounces at ece.cmu.edu] On Behalf Of Sriram Govindan Sent: Wednesday, August 09, 2006 10:25 AM To: haya shem tov Cc: disksim-users at ece.cmu.edu Subject: Re: [Disksim-users] some questions regarding Disksim hi shemtov, Request submission and completion at the IOdriver level is done in handle_new_request and iodriver_access_complete functions respectively in iodriver_disksim.c. From here you can track the complete flow. Hope this helps, sriram.G On Wed, 9 Aug 2006, haya shem tov wrote: > Dear Disksim users, > I am new in disksim and I am trying to integrate disksim in a larger simulator > Using disksim_interface.c. > I am having some difficulties since I do not fully understand the API. > Does anyone has more documentation on disksim interface functions, or more example programs? In detail, in the file syssim_driver.c when the function syssim_report_completion is called , how can I know which request was completed? Is there no ID filed in the Request structure? > Best, > Haya Shemtov > > --------------------------------- > Want to be your own boss? Learn how on Yahoo! Small Business. _______________________________________________ Disksim-users mailing list Disksim-users at ece.cmu.edu https://sos.ece.cmu.edu/mailman/listinfo/disksim-users From zgy04 at 126.com Fri Aug 11 23:22:58 2006 From: zgy04 at 126.com (zgy) Date: Sat, 12 Aug 2006 11:22:58 +0800 Subject: [Disksim-users] When wil the upcoming release of disksim come up?? Message-ID: <200608121116459396746@126.com> Hello,disksim-users? I am using disksim 3.0,it is very good but it can only be compiled with gcc 2.95. I am looking forward to the newer release. Best Luck! ????????zgy ????????zgy04 at 126.com ????????2006-08-12 From xyao at cse.unl.edu Fri Aug 11 23:48:38 2006 From: xyao at cse.unl.edu (Xiaoyu Yao) Date: Fri, 11 Aug 2006 22:48:38 -0500 Subject: [Disksim-users] When wil the upcoming release of disksim come up?? In-Reply-To: <200608121116459396746@126.com> References: <200608121116459396746@126.com> Message-ID: <44DD4F96.6000806@cse.unl.edu> I've run disksim 3.0 with both gcc 2.95 and 3.43. No problem with different gcc versions except some tiny compiling errors due to string formating when using newer version of gcc. But it is not difficult to fix. Xiaoyu zgy wrote: > Hello,disksim-users? > > I am using disksim 3.0,it is very good but it can only be compiled with gcc 2.95. > I am looking forward to the newer release. > > Best Luck! > > ????????zgy > ????????zgy04 at 126.com > ????????2006-08-12 > > ------------------------------------------------------------------------ > > _______________________________________________ > Disksim-users mailing list > Disksim-users at ece.cmu.edu > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > From zgy04 at 126.com Fri Aug 11 23:54:58 2006 From: zgy04 at 126.com (zgy) Date: Sat, 12 Aug 2006 11:54:58 +0800 Subject: [Disksim-users] When wil the upcoming release of disksim comeup?? References: <200608121116459396746@126.com> <44DD4F96.6000806@cse.unl.edu> Message-ID: <200608121154572942320@126.com> Hello,Xiaoyu Yao, thanks for your instant reply. I am simulating a large-scale IO activity,and the x86-32 is too slow for me. Now I am trying to compile disksim on IA64, it seems ok when compiling. But the "segment fault " appears after running. and the gcc version is 3.2.3 Any suggestion? All the best. ????????zgy ????????zgy04 at 126.com ????????2006-08-12 ======= 2006-08-12 11:48:42 Xiaoyu Yao Written: ======= > >I've run disksim 3.0 with both gcc 2.95 and 3.43. No problem with >different gcc versions except some tiny compiling errors due to string >formating when using newer version of gcc. But it is not difficult to fix. > >Xiaoyu > >zgy wrote: >> Hello,disksim-users? >> >> I am using disksim 3.0,it is very good but it can only be compiled with gcc 2.95. >> I am looking forward to the newer release. >> >> Best Luck! >> >> ????????zgy >> ????????zgy04 at 126.com >> ????????2006-08-12 >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Disksim-users mailing list >> Disksim-users at ece.cmu.edu >> https://sos.ece.cmu.edu/mailman/listinfo/disksim-users >> > > = = = = = = = = = = = = = = = = = = = = ?? From xyao at cse.unl.edu Sat Aug 12 01:56:49 2006 From: xyao at cse.unl.edu (Xiaoyu Yao) Date: Sat, 12 Aug 2006 00:56:49 -0500 Subject: [Disksim-users] When wil the upcoming release of disksim comeup?? In-Reply-To: <200608121154572942320@126.com> References: <200608121116459396746@126.com> <44DD4F96.6000806@cse.unl.edu> <200608121154572942320@126.com> Message-ID: <44DD6DA1.1070901@cse.unl.edu> Hi there, The wrong configuration may lead to segment fault. In your case, it is also possible related to the difference between IA32 and IA64. But I don't have environment and experience to run disksim under IA64. I'm just curious about the scale of your simulation and how IA64 can help to improve the simulation speed? Xiaoyu zgy wrote: > Hello,Xiaoyu Yao, > > thanks for your instant reply. I am simulating a large-scale IO activity,and the x86-32 is too slow for me. > Now I am trying to compile disksim on IA64, it seems ok when compiling. But the "segment fault " appears after running. > and the gcc version is 3.2.3 > Any suggestion? > > All the best. > > ????????zgy > ????????zgy04 at 126.com > ????????2006-08-12 > > > ======= 2006-08-12 11:48:42 Xiaoyu Yao Written: ======= > > >> I've run disksim 3.0 with both gcc 2.95 and 3.43. No problem with >> different gcc versions except some tiny compiling errors due to string >> formating when using newer version of gcc. But it is not difficult to fix. >> >> Xiaoyu >> >> zgy wrote: >> >>> Hello,disksim-users? >>> >>> I am using disksim 3.0,it is very good but it can only be compiled with gcc 2.95. >>> I am looking forward to the newer release. >>> >>> Best Luck! >>> >>> ????????zgy >>> ????????zgy04 at 126.com >>> ????????2006-08-12 >>> >>> ------------------------------------------------------------------------ >>> >>> _______________________________________________ >>> Disksim-users mailing list >>> Disksim-users at ece.cmu.edu >>> https://sos.ece.cmu.edu/mailman/listinfo/disksim-users >>> >>> >> > > = = = = = = = = = = = = = = = = = = = = > > > ?? > > From chemical.aniket at gmail.com Mon Aug 14 09:46:03 2006 From: chemical.aniket at gmail.com (Aniket Kulkarni) Date: Mon, 14 Aug 2006 19:16:03 +0530 Subject: [Disksim-users] trace files Message-ID: Where can I get trace files of different block size? If I want to extract such information from my own pc, how can I extract it? aniket -------------- next part -------------- An HTML attachment was scrubbed... URL: From zgy04 at 126.com Mon Aug 14 09:52:51 2006 From: zgy04 at 126.com (Steve) Date: Mon, 14 Aug 2006 21:52:51 +0800 Subject: [Disksim-users] trace files References: Message-ID: <200608142152506047889@126.com> you can get some trace files from here. And the trace has been used in some outstanding papers. http://tesla.hpl.hp.com/public_software/ Where can I get trace files of different block size? If I want to extract such information from my own pc, how can I extract it? aniket -------------- next part -------------- An HTML attachment was scrubbed... URL: From zgy04 at 126.com Mon Aug 14 10:03:05 2006 From: zgy04 at 126.com (Steve) Date: Mon, 14 Aug 2006 22:03:05 +0800 Subject: [Disksim-users] Has anyone succeeded in compiling disksim on IA64? Message-ID: <200608142203050077003@126.com> Hello,disksim-users? I have scratched my head to compile the great disksim 3.0 on an IA64 machine,but when I run disksim on IA64, it give me a segment fault. And I have got to know that the segment fault happened on the 1065 line in the "libparam/util.c" file (i.e. the libparamparse() function call). Because I am confused with the bison toolkit,I can't modify the libparam.y filt to make the libparamparse() usable. Has anyone succeeded in compiling disksim on IA64? Any suggestion or comment is welcome. Best Luck! ????????Steve ????????zgy04 at 126.com ????????2006-08-14 -------------- next part -------------- An HTML attachment was scrubbed... URL: From shahrukh at cs.virginia.edu Tue Aug 15 11:13:00 2006 From: shahrukh at cs.virginia.edu (Shahrukh Rohinton Tarapore) Date: Tue, 15 Aug 2006 11:13:00 -0400 Subject: [Disksim-users] some questions regarding Disksim In-Reply-To: <20060815120056.96065.qmail@web42402.mail.scd.yahoo.com> References: <20060815120056.96065.qmail@web42402.mail.scd.yahoo.com> Message-ID: <44E1E47C.5030106@cs.virginia.edu> the short answer is no, you shouldn't change the ioreq_event struct if you can help it. the ioreq_event struct does have a member called buf which you can use you maintain a pointer to the "system level" request (which contains the ID field). If you look at the disksim_request_arrive function you'll see the system level request parameter is a pointer and that pointer is assigned to the buf member of the ioreq_event. you should do the samething in your implmentation. the reason for not adding the ID field to ioreq_event is that the core disksim structs have been sized to optimize memory allocation during simulation so increasing the size of this struct could unnecessarily slow down your simulation. -- shahrukh haya shem tov wrote: > shahrukh, > your reply was very helpful. > I changed the Request struct in syssim_driver.h to include ID field, > only this did not seem to be enough, since in disksim_interface.c the > Request struct is copied to a ioreq_event struct, so I also added to > this struct the ID field > (does it seem ok?) > this however solved the problem. > thanks, > haya > > */Shahrukh Rohinton Tarapore /* wrote: > > Haya, > The syssim_drive may not be the best example for interfacing a system > level simulator with disksim because syssim_drive only allows one > pending > request at a time, so it is not necessary to identify multiple > requests. > > There is a quick fix to this however. The request structure that > syssim_drive defines can be altered to include a unique id > (probably of > type long int). Then each time you call > disksim_interface_request_arrive > with a new request structure just ensure an unused id is in the > structure. > > When disksim reports completion of a specified request you can > check the > unique id and know exactly which pending request has completed. > > Also realize that if you're going to be putting multiple pending > requests > into disksim the syssim_drive example can not be used. You will > need to > implement your own mechanism for handling multiple requests (via > disksim_interface.c). Hope that helps. > > -- shahrukh > > > On Wed, 9 Aug 2006, haya shem tov wrote: > > > Dear Disksim users, > > I am new in disksim and I am trying to integrate disksim in a > larger simulator > > Using disksim_interface.c. > > I am having some difficulties since I do not fully understand > the API. > > Does anyone has more documentation on disksim interface > functions, or more example programs? In detail, in the file > syssim_driver.c when the function syssim_report_completion is > called , how can I know which request was completed? Is there no > ID filed in the Request structure? > > Best, > > Haya Shemtov > > > > --------------------------------- > > Want to be your own boss? Learn how on Yahoo! Small Business. > > > ------------------------------------------------------------------------ > Stay in the know. Pulse on the new Yahoo.com. Check it out. > From bucy at ece.cmu.edu Tue Aug 15 13:40:41 2006 From: bucy at ece.cmu.edu (John Bucy) Date: Tue, 15 Aug 2006 10:40:41 -0700 Subject: [Disksim-users] some questions regarding Disksim In-Reply-To: <44E1E47C.5030106@cs.virginia.edu> References: <20060815120056.96065.qmail@web42402.mail.scd.yahoo.com> <44E1E47C.5030106@cs.virginia.edu> Message-ID: <4fc1e0430608151040j2acb3a41r5ac52efe94fc7879@mail.gmail.com> Not to mention the fact that (IIRC) there is some code that casts some of those structs around in weird ways that may mysteriously break if you change it. john On 8/15/06, Shahrukh Rohinton Tarapore wrote: > the short answer is no, you shouldn't change the ioreq_event struct if > you can help it. the ioreq_event struct does have a member called buf > which you can use you maintain a pointer to the "system level" request > (which contains the ID field). If you look at the > disksim_request_arrive function you'll see the system level request > parameter is a pointer and that pointer is assigned to the buf member of > the ioreq_event. you should do the samething in your implmentation. > > the reason for not adding the ID field to ioreq_event is that the core > disksim structs have been sized to optimize memory allocation during > simulation so increasing the size of this struct could unnecessarily > slow down your simulation. > > -- shahrukh > > haya shem tov wrote: > > > shahrukh, > > your reply was very helpful. > > I changed the Request struct in syssim_driver.h to include ID field, > > only this did not seem to be enough, since in disksim_interface.c the > > Request struct is copied to a ioreq_event struct, so I also added to > > this struct the ID field > > (does it seem ok?) > > this however solved the problem. > > thanks, > > haya > > > > */Shahrukh Rohinton Tarapore /* wrote: > > > > Haya, > > The syssim_drive may not be the best example for interfacing a system > > level simulator with disksim because syssim_drive only allows one > > pending > > request at a time, so it is not necessary to identify multiple > > requests. > > > > There is a quick fix to this however. The request structure that > > syssim_drive defines can be altered to include a unique id > > (probably of > > type long int). Then each time you call > > disksim_interface_request_arrive > > with a new request structure just ensure an unused id is in the > > structure. > > > > When disksim reports completion of a specified request you can > > check the > > unique id and know exactly which pending request has completed. > > > > Also realize that if you're going to be putting multiple pending > > requests > > into disksim the syssim_drive example can not be used. You will > > need to > > implement your own mechanism for handling multiple requests (via > > disksim_interface.c). Hope that helps. > > > > -- shahrukh > > > > > > On Wed, 9 Aug 2006, haya shem tov wrote: > > > > > Dear Disksim users, > > > I am new in disksim and I am trying to integrate disksim in a > > larger simulator > > > Using disksim_interface.c. > > > I am having some difficulties since I do not fully understand > > the API. > > > Does anyone has more documentation on disksim interface > > functions, or more example programs? In detail, in the file > > syssim_driver.c when the function syssim_report_completion is > > called , how can I know which request was completed? Is there no > > ID filed in the Request structure? > > > Best, > > > Haya Shemtov > > > > > > --------------------------------- > > > Want to be your own boss? Learn how on Yahoo! Small Business. > > > > > > ------------------------------------------------------------------------ > > Stay in the know. Pulse on the new Yahoo.com. Check it out. > > > > _______________________________________________ > Disksim-users mailing list > Disksim-users at ece.cmu.edu > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > > From zgy04 at 126.com Fri Aug 18 07:21:46 2006 From: zgy04 at 126.com (Steve) Date: Fri, 18 Aug 2006 19:21:46 +0800 Subject: [Disksim-users] when the new release of disksim will come out References: <20060815120056.96065.qmail@web42402.mail.scd.yahoo.com> <44E1E47C.5030106@cs.virginia.edu> <4fc1e0430608151040j2acb3a41r5ac52efe94fc7879@mail.gmail.com> Message-ID: <200608181921448021155@126.com> Hello,John Bucy, I wonder when the new release of disksim will come out. At least, I want to know where to get the disk specification file of some newer disk typies whose capacity is more than 10GB. After all, the disk typies described by the disk specification file along wiht disksim 3.0 is too old. All the best. ????????Steve ????????zgy04 at 126.com ????????2006-08-18 ======= 2006-08-16 01:40:54 John Bucy Written: ======= > >Not to mention the fact that (IIRC) there is some code that casts some >of those structs around in weird ways that may mysteriously break if >you change it. > > > >john > >On 8/15/06, Shahrukh Rohinton Tarapore wrote: > > the short answer is no, you shouldn't change the ioreq_event struct if > > you can help it. the ioreq_event struct does have a member called buf > > which you can use you maintain a pointer to the "system level" request > > (which contains the ID field). If you look at the > > disksim_request_arrive function you'll see the system level request > > parameter is a pointer and that pointer is assigned to the buf member of > > the ioreq_event. you should do the samething in your implmentation. > > > > the reason for not adding the ID field to ioreq_event is that the core > > disksim structs have been sized to optimize memory allocation during > > simulation so increasing the size of this struct could unnecessarily > > slow down your simulation. > > > > -- shahrukh > > > > haya shem tov wrote: > > > > > shahrukh, > > > your reply was very helpful. > > > I changed the Request struct in syssim_driver.h to include ID field, > > > only this did not seem to be enough, since in disksim_interface.c the > > > Request struct is copied to a ioreq_event struct, so I also added to > > > this struct the ID field > > > (does it seem ok?) > > > this however solved the problem. > > > thanks, > > > haya > > > > > > */Shahrukh Rohinton Tarapore /* wrote: > > > > > > Haya, > > > The syssim_drive may not be the best example for interfacing a system > > > level simulator with disksim because syssim_drive only allows one > > > pending > > > request at a time, so it is not necessary to identify multiple > > > requests. > > > > > > There is a quick fix to this however. The request structure that > > > syssim_drive defines can be altered to include a unique id > > > (probably of > > > type long int). Then each time you call > > > disksim_interface_request_arrive > > > with a new request structure just ensure an unused id is in the > > > structure. > > > > > > When disksim reports completion of a specified request you can > > > check the > > > unique id and know exactly which pending request has completed. > > > > > > Also realize that if you're going to be putting multiple pending > > > requests > > > into disksim the syssim_drive example can not be used. You will > > > need to > > > implement your own mechanism for handling multiple requests (via > > > disksim_interface.c). Hope that helps. > > > > > > -- shahrukh > > > > > > > > > On Wed, 9 Aug 2006, haya shem tov wrote: > > > > > > > Dear Disksim users, > > > > I am new in disksim and I am trying to integrate disksim in a > > > larger simulator > > > > Using disksim_interface.c. > > > > I am having some difficulties since I do not fully understand > > > the API. > > > > Does anyone has more documentation on disksim interface > > > functions, or more example programs? In detail, in the file > > > syssim_driver.c when the function syssim_report_completion is > > > called , how can I know which request was completed? Is there no > > > ID filed in the Request structure? > > > > Best, > > > > Haya Shemtov > > > > > > > > --------------------------------- > > > > Want to be your own boss? Learn how on Yahoo! Small Business. > > > > > > > > > ------------------------------------------------------------------------ > > > Stay in the know. Pulse on the new Yahoo.com. Check it out. > > > > > > > _______________________________________________ > > Disksim-users mailing list > > Disksim-users at ece.cmu.edu > > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > > > > >_______________________________________________ >Disksim-users mailing list >Disksim-users at ece.cmu.edu >https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > = = = = = = = = = = = = = = = = = = = = ?? -------------- next part -------------- An HTML attachment was scrubbed... URL: From chemical.aniket at gmail.com Thu Aug 24 12:47:41 2006 From: chemical.aniket at gmail.com (Aniket Kulkarni) Date: Thu, 24 Aug 2006 22:17:41 +0530 Subject: [Disksim-users] trace file of the validate format Message-ID: please tell me trace file of validate format i.e. data types of the input parameters. I am waiting for reply. parameters are 1.request flag 2. buffeeraction 3.address (blk no) 4.request size 5.service time 6. next interarrival time plz. waiting for reply. aniket -------------- next part -------------- An HTML attachment was scrubbed... URL: From xyao at cse.unl.edu Thu Aug 24 13:06:09 2006 From: xyao at cse.unl.edu (Xiaoyu Yao) Date: Thu, 24 Aug 2006 12:06:09 -0500 Subject: [Disksim-users] trace file of the validate format In-Reply-To: References: Message-ID: <44EDDC81.3060107@cse.unl.edu> If you seriously read the reply to your post before, you can easily tell the data type. if (sscanf(line, "%c %s %d %d %lf %lf\n", &rw, validate_buffaction, &new->blkno, &new->bcount, &servtime, &validate_nextinter) != 6) Aniket Kulkarni wrote: > please tell me trace file of validate format i.e. data types of the > input parameters. > I am waiting for reply. > parameters are > 1.request flag > 2. buffeeraction > 3.address (blk no) > 4.request size > 5.service time > 6. next interarrival time > plz. waiting for reply. > aniket > ------------------------------------------------------------------------ > > _______________________________________________ > Disksim-users mailing list > Disksim-users at ece.cmu.edu > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > From chemical.aniket at gmail.com Tue Aug 29 09:10:48 2006 From: chemical.aniket at gmail.com (Aniket Kulkarni) Date: Tue, 29 Aug 2006 18:40:48 +0530 Subject: [Disksim-users] where can I get dixtrac? Message-ID: hi, Where can I get source code for dixtrac algorithm? If anybody have that please send me. waiting for reply. aniket -------------- next part -------------- An HTML attachment was scrubbed... URL: From hayashemtov at yahoo.com Tue Aug 29 18:13:17 2006 From: hayashemtov at yahoo.com (haya shem tov) Date: Tue, 29 Aug 2006 15:13:17 -0700 (PDT) Subject: [Disksim-users] Invoking several copies of disksim simultaneously Message-ID: <20060829221317.13625.qmail@web42404.mail.scd.yahoo.com> Dear Disksim users, I intend to use several copies of disksim simultaneously, using disksim_interface.c. I think I can init several disksim copies, even using the same parameter file (?). Then I can activate each disksim copy separately, while they are all synchronized with the system time. The functions syssim_schedule_callback, syssim_deschedule_callback, syssim_report_completion will be invoked only by the disksim copy currently performs an internal event. Am I correct? Is the something I should note? I would also like to join a question previously ask in the forum - Does anybody have disk specification files newer then those published? This could really be a great help, since I feel my research may be much stronger using up-to-date disk specification. Very best, and 10x for any help, Haya Shemtov --------------------------------- Want to be your own boss? Learn how on Yahoo! Small Business. -------------- next part -------------- An HTML attachment was scrubbed... URL: From hayashemtov at yahoo.com Tue Aug 29 18:16:43 2006 From: hayashemtov at yahoo.com (haya shem tov) Date: Tue, 29 Aug 2006 15:16:43 -0700 (PDT) Subject: [Disksim-users] Invoking several copies of disksim simultaneously Message-ID: <20060829221643.59102.qmail@web42410.mail.scd.yahoo.com> Dear Disksim users, I intend to use several copies of disksim simultaneously, using disksim_interface.c. I think I can init several disksim copies, even using the same parameter file (?). Then I can activate each disksim copy separately, while they are all synchronized with the system time. The functions syssim_schedule_callback, syssim_deschedule_callback, syssim_report_completion will be invoked only by the disksim copy currently performs an internal event. Am I correct? Is the something I should note? I would also like to join a question previously ask in the forum - Does anybody have disk specification files newer then those published? This could really be a great help, since I feel my research may be much stronger using up-to-date disk specification. Very best, and 10x for any help, Haya Shemtov --------------------------------- Want to be your own boss? Learn how on Yahoo! Small Business. -------------- next part -------------- An HTML attachment was scrubbed... URL: From bucy at ece.cmu.edu Tue Aug 29 18:20:24 2006 From: bucy at ece.cmu.edu (John Bucy) Date: Tue, 29 Aug 2006 15:20:24 -0700 Subject: Fwd: [Disksim-users] Invoking several copies of disksim simultaneously In-Reply-To: <4fc1e0430608291520n54830fb2we6a2430fd4a1f032@mail.gmail.com> References: <20060829221317.13625.qmail@web42404.mail.scd.yahoo.com> <4fc1e0430608291520n54830fb2we6a2430fd4a1f032@mail.gmail.com> Message-ID: <4fc1e0430608291520jc627aaqc241bce10a19f53a@mail.gmail.com> ---------- Forwarded message ---------- From: John Bucy Date: Aug 29, 2006 3:20 PM Subject: Re: [Disksim-users] Invoking several copies of disksim simultaneously To: haya shem tov Cc: disksim-users at ece.cmu.edu On 8/29/06, haya shem tov wrote: > > Dear Disksim users,I intend to use several copies of disksim simultaneously, > using disksim_interface.c. I think I can init several disksim copies, > even using the same parameter file (?). > Then I can activate each disksim copy separately, > while they are all synchronized with the system time. > The functions syssim_schedule_callback, syssim_deschedule_callback, > syssim_report_completion will be invoked only by the disksim copy > currently performs an internal event. > Am I correct? Is the something I should note? I would be shocked if this actually worked correctly; there is a huge amount of global state inside disksim; multiple instances will corrupt each other. If you really need to do this, you can probably fork off processes and do the synchronization with some kind of ipc. > I would also like to join a question previously ask in the forum - > Does anybody have disk specification files newer then those published? > This could really be a great help, since I feel my research may be much > stronger using up-to-date disk specification. I've been working on this in my (limited) spare time and I wish it was done 9 months ago but there are only so many hours in the day :( john From xyao at cse.unl.edu Tue Aug 29 19:38:24 2006 From: xyao at cse.unl.edu (Xiaoyu Yao) Date: Tue, 29 Aug 2006 18:38:24 -0500 Subject: Fwd: [Disksim-users] Invoking several copies of disksim simultaneously In-Reply-To: <4fc1e0430608291520jc627aaqc241bce10a19f53a@mail.gmail.com> References: <20060829221317.13625.qmail@web42404.mail.scd.yahoo.com> <4fc1e0430608291520n54830fb2we6a2430fd4a1f032@mail.gmail.com> <4fc1e0430608291520jc627aaqc241bce10a19f53a@mail.gmail.com> Message-ID: <44F4CFF0.3000909@cse.unl.edu> Can you open an account for public CVS access to the latest source so that more developers interested in disksim can learn and contribute? Xiaoyu John Bucy wrote: > ---------- Forwarded message ---------- > From: John Bucy > Date: Aug 29, 2006 3:20 PM > Subject: Re: [Disksim-users] Invoking several copies of disksim > simultaneously > To: haya shem tov > Cc: disksim-users at ece.cmu.edu > > > On 8/29/06, haya shem tov wrote: >> >> Dear Disksim users,I intend to use several copies of disksim >> simultaneously, >> using disksim_interface.c. I think I can init several disksim copies, >> even using the same parameter file (?). >> Then I can activate each disksim copy separately, >> while they are all synchronized with the system time. >> The functions syssim_schedule_callback, syssim_deschedule_callback, >> syssim_report_completion will be invoked only by the disksim copy >> currently performs an internal event. >> Am I correct? Is the something I should note? > > I would be shocked if this actually worked correctly; there is a huge > amount of global state inside disksim; multiple instances will corrupt > each other. If you really need to do this, you can probably fork off > processes and do the synchronization with some kind of ipc. > > >> I would also like to join a question previously ask in the forum - >> Does anybody have disk specification files newer then those published? >> This could really be a great help, since I feel my research may be much >> stronger using up-to-date disk specification. > > I've been working on this in my (limited) spare time and I wish it was > done 9 months ago but there are only so many hours in the day :( > > > > john > _______________________________________________ > Disksim-users mailing list > Disksim-users at ece.cmu.edu > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users >