On Mon, Jan 29, 2018 at 11:40:27AM +0200, Avi Kivity wrote: > > > On 01/25/2018 03:08 PM, Brian Foster wrote: > > On Thu, Jan 25, 2018 at 10:50:40AM +0200, Avi Kivity wrote: > > > On 01/23/2018 07:39 PM, Brian Foster wrote: > > > > Yeah, could be.. perhaps the issue is that despite the large amount of > > > > total free space, the free space is too fragmented to satisfy a > > > > particular allocation request..? > > > from to extents blocks pct > > > 1 1 2702 2702 0.00 > > > 2 3 690 1547 0.00 > > > 4 7 115 568 0.00 > > > 8 15 60 634 0.00 > > > 16 31 63 1457 0.00 > > > 32 63 102 4751 0.00 > > > 64 127 7940 895365 0.19 > > > 128 255 49680 12422100 2.67 > > > 256 511 1025 417078 0.09 > > > 512 1023 4170 3660771 0.79 > > > 1024 2047 2168 3503054 0.75 > > > 2048 4095 2567 7729442 1.66 > > > 4096 8191 8688 59394413 12.76 > > > 8192 16383 310 3100186 0.67 > > > 16384 32767 112 2339935 0.50 > > > 32768 65535 35 1381122 0.30 > > > 65536 131071 8 651391 0.14 > > > 131072 262143 2 344196 0.07 > > > 524288 1048575 4 2909925 0.62 > > > 1048576 2097151 3 3550680 0.76 > > > 4194304 8388607 10 82497658 17.72 > > > 8388608 16777215 10 158022653 33.94 > > > 16777216 24567552 5 122778062 26.37 > > > total free extents 80469 > > > total free blocks 465609690 > > > average free extent size 5786.2 > > > > > > Looks like plenty of free large extents, with most of the free space > > > completely, unfragmented. > > > > > Indeed.. You need to look at each AG, not the overall summary. You could have a suboptimal AG hidden in amongst that (e.g. near ENOSPC) and it's that one AG that is causing all your problems. There's many reasons this can happen, but the most common is the working files in a directory (or subset of directories in the same AG) have a combined space usage of larger than an AG .... > > > Lots of 16MB-32MB extents, too. 32MB is our allocation hint size, could have > > > something to do with it. > > > > > Most likely. Based on this, it's hard to say for certain why you'd be > > running into allocation latency caused by busy extents. One of only two reasons: 1. the AG has a large enough free space, but they are all marked busy (i.e. just been freed), or 2. The extent selected has had the busy range trimmed out of it and it's now less than the minimum extent length requested. Both cases imply that we're allocating extents that have been very recently freed, and that implies there is no other suitable non-busy free space in the AG. Hence the need to look at the per-AG freespace pattern rather than the global summary. Also, it's worth dumping the freespace via xfs_spaceman as it walks the in memory trees rather than the on-disk trees and so is properly coherent with operations in progress. (i.e. xfs_spaceman -c "freesp ..." /mntpt) Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
