Re: Low-level trace-buffers in CAM

-Original Message- From: John Baldwin Date: 2015-11-24, Tuesday at 18:07 To: "freebsd-hack...@freebsd.org" Cc: Adrian Chadd , Ravi Pokala , "freebsd-geom@freebsd.org"

Re: Low-level trace-buffers in CAM

eebsd.org>, "freebsd-hack...@freebsd.org" <freebsd-hack...@freebsd.org>, "k...@freebsd.org" <k...@freebsd.org>, "i...@freebsd.org" <i...@freebsd.org>, "sco...@freebsd.org" <sco...@freebsd.org> Subject: Re:

Re: Low-level trace-buffers in CAM

-Original Message- From: Ravi Pokala Date: 2015-10-26, Monday at 21:21 To: Adrian Chadd Cc: "freebsd-geom@freebsd.org" , "freebsd-s...@freebsd.org" , "freebsd-hack...@freebsd.org"

Low-level trace-buffers in CAM

Hi folks, This is an updated re-send of a message I originally sent about a year ago, during MeetBSD 2014. A few people expressed interest in person, but no one ever followed up on the lists. I'm bringing this up again, in the

Re: Low-level trace-buffers in CAM

-Original Message- From: Scott Long Date: 2015-10-26, Monday at 20:08 To: Ravi Pokala Cc: "freebsd-geom@freebsd.org" , "freebsd-s...@freebsd.org" , "freebsd-hack...@freebsd.org"

Re: Converting LBAs to byte offsets through the GEOM stack

Hi Adrian, So I had a thing that I attached commands to that would funnel down to the geom layer that did this mirroring/caching/remapping thing, and it would handle schedule the commands to whatever block(s) on whatever disk(s) actually represented that particular logical offset. I actually had

Converting LBAs to byte offsets through the GEOM stack

Hi folks, When you issue a BIO, the requested byte offset (bio_offset) gets transformed by each layer of the GEOM stack as needed. If the bottom of the stack is a physical disk, g_disk_start() transforms the final offset to a device block address (bio_pblkno), which the disk device driver uses as

Re: Converting LBAs to byte offsets through the GEOM stack

I've only done this manually... It isn't too hard, as all the partitioning schemes are simple offsets, and the stripe should be regular... The *partitioning schemes*, yes. But once you start building up more layers, it gets complicated. The issue w/ manually mapping, is that you might loose a

Low-level trace-buffers in CAM

Hi folks, At Panasas, we found it useful to have a trace buffer for each ATA drive in the system, for gathering information about command sequences and latency. I implemented that functionality for our old 7-ish ATA driver, and it works quite well, with fairly low overhead. For example (sorry for