On 5/11/23 20:41, Ming Lin wrote:
Hi list,
I have an application that needs to use buffered_io to access SMR disk
for good performance.
From "ZBD Support Restrictions" at https://zonedstorage.io/docs/linux/overview
"
Direct IO Writes The kernel page cache does not guarantee that cached
dirty pages will be flushed to a block device in sequential sector
order. This can lead to unaligned write errors if an application uses
buffered writes to write to the sequential write required zones of a
device. To avoid this pitfall, applications that directly use a zoned
block device without a file system should always use direct I/O
operations to write to the sequential write required zones of a
host-managed disk (that is, they should issue write() system calls
with a block device "file open" that uses the O_DIRECT flag).
"
Raw zbd disk only supports direct_io.
Does dm-zoned support buffered io (without O_DIRECT)?
Yes. But I _think_ the above paragraph is ever so slightly outdated, as
we've spent quite a lot of time fixing sequential writes (cf blk-zoned
etc). So while dm-zoned is using bufferet writes there won't be any
sequential write issues.
At least, I have not uncovered any of those during testing.
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@xxxxxxx +49 911 74053 688
SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), Geschäftsführer: Ivo Totev, Andrew
Myers, Andrew McDonald, Martje Boudien Moerman
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://listman.redhat.com/mailman/listinfo/dm-devel
