From rsalarifardce87 at gmail.com Sat Nov 2 01:40:46 2013 From: rsalarifardce87 at gmail.com (Raziye Salarifard) Date: Sat, 2 Nov 2013 09:10:46 +0330 Subject: [Disksim-users] hybrid cache for disksim Message-ID: Hi, I wanna design a hybrid cache for disksim, this cache includes SSD and HDD, at the frist I extended disksim with SSD microsoft extension,but I have no idea about how to design the cache with these available SSD and HDD ? Regards -- *Raziye Salarifard* *M.S. Student* *Computer Engineering Department,Sharif University of Technology* -------------- next part -------------- An HTML attachment was scrubbed... URL: From salinezhad at gmail.com Tue Nov 19 01:40:21 2013 From: salinezhad at gmail.com (saeedeh alinezhad) Date: Tue, 19 Nov 2013 10:10:21 +0330 Subject: [Disksim-users] RAID in Flashsim? Message-ID: hello everyone I wonder if it is possible to add more than one SSD in the flashsim to create a RAID structure? is it code support such modification??? is it possible to use the RAID of disksim and the FTL of flashsim? I mean are their codes match with each other and portable?? -- Regards, Saeideh Alinezhad -------------- next part -------------- An HTML attachment was scrubbed... URL: From salinezhad at gmail.com Tue Nov 19 01:42:34 2013 From: salinezhad at gmail.com (saeedeh alinezhad) Date: Tue, 19 Nov 2013 10:12:34 +0330 Subject: [Disksim-users] Flashsim in Disksim4? Message-ID: hello every one flashsim is add to the disksim3 and work with it. is there anyone who install it on disksim4 and compile it correctly?is it possible? -- Regards, Saeideh Alinezhad -------------- next part -------------- An HTML attachment was scrubbed... URL: From tiratatp at cs.uchicago.edu Tue Nov 19 23:59:51 2013 From: tiratatp at cs.uchicago.edu (Tiratat Patana-anake) Date: Tue, 19 Nov 2013 22:59:51 -0600 Subject: [Disksim-users] Disable Write Cache in DiskSim 4.0 Message-ID: Hi, I'm trying to disable a "write cache" in DiskSim 4.0. I'm not sure which parameters in parv file that control this. The closest I can find is "Fast write level" which I can set to level 0, 1 and 2, and 0 should mean disable write cache in all cases. Is this true? Is there a better way to disable write cache? Second, when I set "Fast write level" to 0, the disksim only ran some of the trace. (e.g. in a trace that have 110892 I/Os, disksim only ran 5 I/Os) Thank you. Best regards Knot -- Tiratat (Knot) Patana-anake ?????? (????) ????????? From tiratatp at cs.uchicago.edu Wed Nov 20 00:12:31 2013 From: tiratatp at cs.uchicago.edu (Tiratat Patana-anake) Date: Tue, 19 Nov 2013 23:12:31 -0600 Subject: [Disksim-users] assertion failed: in disk_buffer_sector_done() (disksim_diskctlr.c:4360): currdisk->fpcheck: Message-ID: Hi Have anyone ran into this problem, assertion failed: in disk_buffer_sector_done() (disksim_diskctlr.c:4360): currdisk->fpcheck: Apparently, these is this comment in the header (disksim_disk.h), // This is a forward progress check. There have been a number of // bugs over the years that cause the controller to get into an // infinite loop, reading the same sector or track over and over. // Every time we do a reposition, this is reset (currently // statically, should be set to the track length) and // disk_buffer_sector_done() decrements it and asserts that it // hasn't gotten to 0. 2 questions, 1. any workaround of this bug? 2. does anyone know what is this bug? Best regards Knot -- Tiratat (Knot) Patana-anake ?????? (????) ????????? From 815823554 at qq.com Wed Nov 20 09:12:35 2013 From: 815823554 at qq.com (=?gb18030?B?yM7O0tDQ?=) Date: Wed, 20 Nov 2013 22:12:35 +0800 Subject: [Disksim-users] A question about new trace Message-ID: Hi,everyone! I download a new trace Financial1.spc on the internet. And I made some changes of the functions just as the disksim manual says. I run the order " ../src/disksim ascii.parv spc.outv spc Financial1.spc 0 " in the terminal. However, the results is as follows: unexpected requests location:devno13, blkno 60364, bcout 8192 Assertion failed: Simtime =146.977584 totalreqs =4357 disksim:disksim_logorg.c:750:logorg_maprequest:Assertion 'logorgno !=-1' failed. Aborted What should I do to deal with it? -------------- next part -------------- An HTML attachment was scrubbed... URL: From tiratatp at cs.uchicago.edu Wed Nov 20 09:24:47 2013 From: tiratatp at cs.uchicago.edu (Tiratat Patana-anake) Date: Wed, 20 Nov 2013 08:24:47 -0600 Subject: [Disksim-users] A question about new trace In-Reply-To: References: Message-ID: This is because disksim cannot find where to write your IO. This might happen because your disk (as configured in parv file) is too small or doesn't match topology of your trace file (e.g. trace file used RAID but parv don't). -- Tiratat (Knot) Patana-anake ?????? (????) ????????? On Wed, Nov 20, 2013 at 8:12 AM, ??? <815823554 at qq.com> wrote: > Hi,everyone! > I download a new trace Financial1.spc on the internet. And I made some > changes of the functions just as the disksim manual says. I run the order " > ../src/disksim ascii.parv spc.outv spc Financial1.spc 0 " in the terminal. > However, the results is as follows: > > unexpected requests location:devno13, blkno 60364, bcout 8192 > > Assertion failed: > Simtime =146.977584 > totalreqs =4357 > disksim:disksim_logorg.c:750:logorg_maprequest:Assertion 'logorgno !=-1' > failed. > Aborted > > What should I do to deal with it? > > > > _______________________________________________ > Disksim-users mailing list > Disksim-users at ece.cmu.edu > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > From tiratatp at cs.uchicago.edu Wed Nov 20 10:56:29 2013 From: tiratatp at cs.uchicago.edu (Tiratat Patana-anake) Date: Wed, 20 Nov 2013 09:56:29 -0600 Subject: [Disksim-users] assertion failed: in disk_buffer_sector_done() (disksim_diskctlr.c:4360): currdisk->fpcheck: In-Reply-To: References: Message-ID: Solution!!! Today I found that the culprit (at least in my case) is that one of the IO's block count is greater than 6000. (6000 is the *hardcoded* number of fpcheck) So I changed to 65537 (disksim_diskctlr.c line 1920) and it was solved! (at least for now) Have you ever ran into this problem? Any comment on my dirty workaround? -- Tiratat (Knot) Patana-anake ?????? (????) ????????? On Tue, Nov 19, 2013 at 11:12 PM, Tiratat Patana-anake wrote: > Hi > > Have anyone ran into this problem, > assertion failed: in disk_buffer_sector_done() > (disksim_diskctlr.c:4360): currdisk->fpcheck: > > Apparently, these is this comment in the header (disksim_disk.h), > // This is a forward progress check. There have been a number of > // bugs over the years that cause the controller to get into an > // infinite loop, reading the same sector or track over and over. > // Every time we do a reposition, this is reset (currently > // statically, should be set to the track length) and > // disk_buffer_sector_done() decrements it and asserts that it > // hasn't gotten to 0. > > 2 questions, > 1. any workaround of this bug? > 2. does anyone know what is this bug? > > Best regards > Knot > > -- > Tiratat (Knot) Patana-anake > ?????? (????) ????????? From tiratatp at cs.uchicago.edu Thu Nov 21 13:36:04 2013 From: tiratatp at cs.uchicago.edu (Tiratat Patana-anake) Date: Thu, 21 Nov 2013 12:36:04 -0600 Subject: [Disksim-users] Bandwidth reporting Message-ID: Hi Is there a way to get average bandwidth information from the simulation? Best regards Knot -- Tiratat (Knot) Patana-anake ?????? (????) ????????? From mujtaba.tarihi at gmail.com Sat Nov 23 13:53:48 2013 From: mujtaba.tarihi at gmail.com (Mujtaba Tarihi) Date: Sat, 23 Nov 2013 22:23:48 +0330 Subject: [Disksim-users] assertion failed: in disk_buffer_sector_done() (disksim_diskctlr.c:4360): currdisk->fpcheck: In-Reply-To: References: Message-ID: Hello >From what I gathered from discussions on this mailing list, this counter is used to detect infinite loops and this is the way to deal with it when you are running really long traces. On Wed, Nov 20, 2013 at 7:26 PM, Tiratat Patana-anake < tiratatp at cs.uchicago.edu> wrote: > Solution!!! > > Today I found that the culprit (at least in my case) is that one of > the IO's block count is greater than 6000. > (6000 is the *hardcoded* number of fpcheck) > So I changed to 65537 (disksim_diskctlr.c line 1920) and it was > solved! (at least for now) > > Have you ever ran into this problem? > Any comment on my dirty workaround? > -- > Tiratat (Knot) Patana-anake > ?????? (????) ????????? > > > On Tue, Nov 19, 2013 at 11:12 PM, Tiratat Patana-anake > wrote: > > Hi > > > > Have anyone ran into this problem, > > assertion failed: in disk_buffer_sector_done() > > (disksim_diskctlr.c:4360): currdisk->fpcheck: > > > > Apparently, these is this comment in the header (disksim_disk.h), > > // This is a forward progress check. There have been a number of > > // bugs over the years that cause the controller to get into an > > // infinite loop, reading the same sector or track over and over. > > // Every time we do a reposition, this is reset (currently > > // statically, should be set to the track length) and > > // disk_buffer_sector_done() decrements it and asserts that it > > // hasn't gotten to 0. > > > > 2 questions, > > 1. any workaround of this bug? > > 2. does anyone know what is this bug? > > > > Best regards > > Knot > > > > -- > > Tiratat (Knot) Patana-anake > > ?????? (????) ????????? > _______________________________________________ > 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: