From eestolan at soe.ucsc.edu Tue Dec 2 19:57:42 2008 From: eestolan at soe.ucsc.edu (Esteban Molina-Estolano) Date: Tue, 2 Dec 2008 16:57:42 -0800 Subject: [Disksim-users] DiskSim under a Python simulator? Message-ID: I would like to run multiple concurrent instances of DiskSim as slaves to a simulator written in Python. I'm thinking about writing a Python extension module for disksim_interface.h/c, but I'm hoping there's an easier way. Has anyone already created a Python extension module for DiskSim that I might use? I looked around, but didn't find one. Alternately, does someone know of a straightforward way to drive DiskSim from Python without an extension module? Thanks! ~ Esteban From lucky.wu.han at gmail.com Fri Dec 5 09:58:29 2008 From: lucky.wu.han at gmail.com (lucky.wu.han) Date: Fri, 5 Dec 2008 22:58:29 +0800 Subject: [Disksim-users] Why most request don't be handled? Message-ID: <200812052255205150970@gmail.com> hi,all I used a kind of trace which's request number is 567894(all of they are read operation). I used disksim4.0 to simulate this trace run in a storage system. The report of the disksim say that "overall I/O system total requests handled: 375522". It is very surprising! The trace request number is 567894. Why request number unequal to total requests handled ? Is it because the request is too intensive, disksim discard some requests? or maybe is other reasons? Thank you! 2008-12-05 lucky.wu.han -------------- next part -------------- An HTML attachment was scrubbed... URL: From gunusha at gmail.com Fri Dec 5 10:18:41 2008 From: gunusha at gmail.com (vilo gunusha) Date: Fri, 5 Dec 2008 23:18:41 +0800 Subject: [Disksim-users] disksim_diskctlr.c:4364 Message-ID: <200812052318199531583@gmail.com> hi,all I use a trace which involve 6 devno. I find a strange phenomena. I run all requests except for which's devno is 2 in disksim4.0 successfully. When i use the requests which's devno is 2 in disksim, it print error: "assertion failed:in disk_buffer_sector_done()(disksim_diskctlr.c:4364):currdisk->fpcheck:" I find the requests with devno 2 have no difference between requests with devno 0,1,3,4,5. Who have encountered this problem? What the reason is it? 2008-12-05 vilo gunusha -------------- next part -------------- An HTML attachment was scrubbed... URL: From guyh at iis.u-tokyo.ac.jp Sat Dec 20 13:03:43 2008 From: guyh at iis.u-tokyo.ac.jp (Yanhui-iis) Date: Sun, 21 Dec 2008 03:03:43 +0900 Subject: [Disksim-users] about how to evaluate the energy consumption Message-ID: <000001c962cd$4d5c92a0$e815b7e0$@u-tokyo.ac.jp> Dear all, I met some questions. When I simulated a storage system, can I use DiskSim to evaluate the energy consumption? If so, how to do? What are the energy-related metrics in DiskSim? Thanks a lot. Regards, Yanhui-IIS From zhengwj07 at mails.tsinghua.edu.cn Tue Dec 16 03:41:50 2008 From: zhengwj07 at mails.tsinghua.edu.cn (zhengwj07) Date: Tue, 16 Dec 2008 16:41:50 +0800 Subject: [Disksim-users] help about SSD model with cache support Message-ID: <200812161641500464995@mails.tsinghua.edu.cn> Hi, the SSD model comes with no support of cache simulation, but I want to simulate a SSD based storage system with cache support in the controller or in the iodriver. I tried to run SSD with a type=3 controller including a cachemem model, but the following error comes up: Stopping simulation because of saturation: simtime 136334.931000, totalreqs 10000 last request: dev=0, blk=44403360, cnt=8, 1 (R==1) The parameter file, output file and trace file are in the attachmemt. I want to know whether there is a solution to it or whether there is another way to simulate a SSD with cache support?? Thx! Looking foward for the reply! 2008-12-16 zhengwj07 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: ssdcache.parv Type: application/octet-stream Size: 9882 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: out.outv Type: application/octet-stream Size: 149232 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: postmark.trace Type: application/octet-stream Size: 1675549 bytes Desc: not available URL: From yueyinliang at gmail.com Sat Dec 20 20:15:40 2008 From: yueyinliang at gmail.com (Yinliang Yue) Date: Sun, 21 Dec 2008 09:15:40 +0800 Subject: [Disksim-users] help about SSD model with cache support In-Reply-To: <200812161641500464995@mails.tsinghua.edu.cn> References: <200812161641500464995@mails.tsinghua.edu.cn> Message-ID: I think you should reduce the I/O intensiveness of the workload and try again. Regards! Yinliang Yue HUST, Wuhan, China On 12/16/08, zhengwj07 wrote: > > Hi, > the SSD model comes with no support of cache simulation, but I want to > simulate a SSD based storage system with cache support in the controller or > in the iodriver. I tried to run SSD with a type=3 controller including a > cachemem model, but the following error comes up: > > Stopping simulation because of saturation: simtime 136334.931000, totalreqs > 10000 > last request: dev=0, blk=44403360, cnt=8, 1 (R==1) > > The parameter file, output file and trace file are in the attachmemt. > I want to know whether there is a solution to it or whether there is > another way to simulate a SSD with cache support?? > > Thx! Looking foward for the reply! > > 2008-12-16 > ------------------------------ > zhengwj07 > > _______________________________________________ > Disksim-users mailing list > Disksim-users at ece.cmu.edu > https://sos.ece.cmu.edu/mailman/listinfo/disksim-users > > > -- Best regards! Yinliang YUE HUST, Wuhan, China yueyinliang at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: