On 1/31/18 7:36 AM, Anand Jain wrote: > > > On 01/31/2018 09:42 PM, Nikolay Borisov wrote: > > >>>> So usually this should be functionality handled by the raid/san >>>> controller I guess, > but given that btrfs is playing the role of a >>>> controller here at what point are we drawing the line of not >>>> implementing block-level functionality into the filesystem ? >>> >>> Don't worry this is not invading into the block layer. How >>> can you even build this functionality in the block layer ? >>> Block layer even won't know that disks are mirrored. RAID >>> does or BTRFS in our case. >>> >> >> By block layer I guess I meant the storage driver of a particular raid >> card. Because what is currently happening is re-implementing >> functionality that will generally sit in the driver. So my question was >> more generic and high-level - at what point do we draw the line of >> implementing feature that are generally implemented in hardware devices >> (be it their drivers or firmware). > > Not all HW configs use RAID capable HBAs. A server connected to a SATA > JBOD using a SATA HBA without MD will relay on BTRFS to provide all the > features and capabilities that otherwise would have provided by such a > presumable HW config.
That does sort of sound like means implementing some portion of the HBA features/capabilities in the filesystem. To me it seems this this could be workable at the fs level, provided it deals just with policies and remains hardware-neutral. However most of the use cases appear to involve some hardware-dependent knowledge or assumptions. What happens when someone sets this on a virtual disk, or say a (persistent) memory-backed block device? Case #6 seems to open up some potential for unexpected interactions (which may be hard to reproduce, esp. in error/recovery scenarios). Case #2 takes a devid, but I notice btrfs_device::devid says, "the internal btrfs device id". How does a user obtain that internal value so it can be set as a mount option? Thanks, Ed >>>>> :: >>>>>>> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c >>>>>>> index 39ba59832f38..478623e6e074 100644 >>>>>>> --- a/fs/btrfs/volumes.c >>>>>>> +++ b/fs/btrfs/volumes.c >>>>>>> @@ -5270,6 +5270,16 @@ static int find_live_mirror(struct >>>>>>> btrfs_fs_info *fs_info, >>>>>>> num = map->num_stripes; >>>>>>> switch(fs_info->read_mirror_policy) { >>>>>>> + case BTRFS_READ_MIRROR_BY_DEV: >>>>>>> + optimal = first; >>>>>>> + if (test_bit(BTRFS_DEV_STATE_READ_MIRROR, >>>>>>> + &map->stripes[optimal].dev->dev_state)) >>>>>>> + break; >>>>>>> + if (test_bit(BTRFS_DEV_STATE_READ_MIRROR, >>>>>>> + &map->stripes[++optimal].dev->dev_state)) >>>>>>> + break; >>>>>>> + optimal = first; >>>>>> >>>>>> you set optimal 2 times, the second one seems redundant. >>>>> >>>>> No actually. When both the disks containing the stripe does not >>>>> have the BTRFS_DEV_STATE_READ_MIRROR, then I would just want to >>>>> use first found stripe. >>>> >>>> Yes, and the fact that you've already set optimal = first right after >>>> BTRFS_READ_MIRROR_BY_DEV ensures that, no ? Why do you need to again >>>> set >>>> optimal right before the final break? What am I missing here? >>> >>> Ah. I think you are missing ++optimal in the 2nd if. >> >> You are right, but I'd prefer you index the stripes array with 'optimal' >> and 'optimal + 1' and leave just a single assignment > > Ok. Will improve that. > > Thanks, Anand > > >>> >>> Thanks, Anand >>> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in >> the body of a message to majord...@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > -- > To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in > the body of a message to majord...@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html