Jens, I'm very appreciative that you and others are giving this time and consideration.
For the replay host the OS is RHEL6.3. (The originating host is RHEL6.4.) [root@x3650M4 cache_stats]# uname -r 2.6.32-279.el6.x86_64 I've completed a new trace capture and replay using a single device. The originating block trace encountered no dropped events this time. I used fio 2.1.4 this time. The replay once again stopped prematurely. Here is the synopsis based on analysis of the block trace captured on the originating host to the block trace captured on the replay host. I have another email to this thread with all the details that I will send upon request. I would postulate since others have had success with replay that the issue is with something I am doing vs a fio issue. Just need help identifying what I need to do differently. That is to say that running a different version of fio will likely result in the same issue unless there are fixes in replay that map to what has been observed here. I rather imagine that discovering my issue will involve someone well versed in replay taking time to review my procedure and results (in the detailed email not yet sent). Comparative analysis replay to original Capture time .88% % reads +1.16% % writes -1.16% Block range 104% Queue operations .78% Blocks queued .78% % sequential queued 108% Max blocks per sequential run 31.4% Completion operations 1.42% Blocks completed .78% % sequential completions 110% Eric -----Original Message----- From: Jens Axboe [mailto:[email protected]] Sent: Thursday, February 27, 2014 12:49 PM To: Lamkin, Eric; David Nellans Cc: [email protected] Subject: Re: fio replay On 2014-02-26 13:55, Lamkin, Eric wrote: > Jens, > > Sorry I don't know how to compile. > Hope that doesn't mean no one will ever replay to me again. It's pretty easy, basically boils down to doing: $ git clone git://git.kernel.dk/fio $ cd fio $ make and then run ./fio. What OS are you using? -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
