On 07/29/2016 04:20 PM, J. Bruce Fields wrote: > On Fri, Jul 29, 2016 at 03:40:00PM -0400, Anna Schumaker wrote: >> On 07/29/2016 02:59 PM, J. Bruce Fields wrote: >>> On Fri, May 13, 2016 at 04:58:06PM -0400, Anna Schumaker wrote: >>>> On 05/13/2016 04:31 PM, J. Bruce Fields wrote: >>>>> On Sun, May 01, 2016 at 10:37:33AM -0700, Christoph Hellwig wrote: >>>>>> I might sound like a broken record, but I'd feel much happier if this >>>>>> had extensive xfstests coverage. Xfstests has over one hundred tests for >>>>>> file clones, and many of them should be easily adapatable. >>>>> >>>>> Anna, have you looked at this yet? >>>> >>>> Yep! I just sent out what I came up with :) >>> >>> Sorry for the lack of response. For some reason I don't seem to have >>> the updated version in my mailboxes. Do you have a more recent version? >> >> I'm not sure, so I'll make sure my code still works and then resubmit! >> >>> >>>>> I don't see any obvious problem with the nfsd code, other than the >>>>> obvious issue with large synchronous copies tying up server threads and >>>>> leaving clients waiting--but maybe we should just see how people end up >>>>> using it and deal with the problems as they come up. >>> >>> I'm still worrying about this, though. >>> >>> As a simple stopgap, could we just set *some* maximum on the size of the >>> copy? Or better yet on the time?--that'd let filesystems with >>> clone-like features copy the whole file without blocking an nfsd thread >>> indefinitely in the case of other filesystems. >> >> Would there be a good way of figuring out the time a copy would take? > > Can we set some sort of timer to signal our thread after a limit? Then > hopefully the copy loop gets interrupted and we can return the amount > copied so far. (And hopefully the client has actually set the > contiguous flag so it can continue where it left off.) > >> Capping with an arbitrary size would definitely be simpler, so I'll >> look into adding that. > > I'm not sure how to set the limit. The downside (assuming the > client/application handle the short copy correctly) is that data can > stop flowing while we wait for the client to send us the next copy, but > I'm not sure how high the cap needs to be before that becomes > negligible. I've been playing around with copy caps and I found that a 4MB cap has performance pretty close to no cap at all. Here is my performance data (averages across several runs): |-------|---------|---------|---------|---------|---------|---------|---------| |NFSv4.1| 512 MB | 1024 MB | 1536 MB | 2048 MB | 2560 MB | 3072 MB | 5120 MB | |-------|---------|---------|---------|---------|---------|---------|---------| |user | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | |system | 0.28s | 0.48s | 0.72s | 0.94s | 1.17s | 1.42s | 2.37s | |cpu | 18% | 14% | 14% | 14% | 14% | 13% | 13% | |total | 1.547s | 3.269s | 5.040s | 6.690s | 8.363s | 10.146s | 16.729s | |-------|---------|---------|---------|---------|---------|---------|---------| |read | 4096 | 8192 | 12288 | 16384 | 20480 | 24576 | 40960 | |write | 4098 | 8203 | 12302 | 16402 | 20497 | 24668 | 40996 | |commit | 9 | 18 | 27 | 36 | 46 | 55 | 93 | |-------|---------|---------|---------|---------|---------|---------|---------| |-------|---------|---------|---------|---------|---------|---------|---------| |No Cap | 512 MB | 1024 MB | 1536 MB | 2048 MB | 2560 MB | 3072 MB | 5120 MB | |-------|---------|---------|---------|---------|---------|---------|---------| |user | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | |system | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | |cpu | 0% | 0% | 0% | 0% | 0% | 0% | 0% | |total | 1.476s | 3.068s | 4.569s | 6.123s | 7.701s | 9.265s | 15.438s | |-------|---------|---------|---------|---------|---------|---------|---------| |copy | 1 | 1 | 1 | 2 | 2 | 2 | 3 | |commit | 1 | 1 | 1 | 2 | 2 | 2 | 3 | |-------|---------|---------|---------|---------|---------|---------|---------| |-------|---------|---------|---------|---------|---------|---------|---------| |1MB Cap| 512 MB | 1024 MB | 1536 MB | 2048 MB | 2560 MB | 3072 MB | 5120 MB | |-------|---------|---------|---------|---------|---------|---------|---------| |user | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | |system | 0.01s | 0.02s | 0.03s | 0.04s | 0.05s | 0.07s | 0.10s | |cpu | 0% | 0% | 0% | 0% | 0% | 0% | 0% | |total | 1.659s | 3.118s | 4.930s | 6.647s | 8.110s | 9.637s | 16.030s | |-------|---------|---------|---------|---------|---------|---------|---------| |copy | 512 | 1024 | 1536 | 2048 | 2560 | 3072 | 5120 | |commit | 512 | 1024 | 1536 | 2048 | 2560 | 3072 | 5120 | |-------|---------|---------|---------|---------|---------|---------|---------| |-------|---------|---------|---------|---------|---------|---------|---------| |2MB Cap| 512 MB | 1024 MB | 1536 MB | 2048 MB | 2560 MB | 3072 MB | 5120 MB | |-------|---------|---------|---------|---------|---------|---------|---------| |user | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | |system | 0.01s | 0.01s | 0.02s | 0.03s | 0.03s | 0.04s | 0.07s | |cpu | 0% | 0% | 0% | 0% | 0% | 0% | 0% | |total | 1.757s | 3.351s | 4.791s | 6.522s | 8.090s | 9.594s | 15.977s | |-------|---------|---------|---------|---------|---------|---------|---------| |copy | 256 | 512 | 768 | 1024 | 1280 | 1536 | 2560 | |commit | 256 | 512 | 768 | 1024 | 1280 | 1536 | 2560 | |-------|---------|---------|---------|---------|---------|---------|---------| |-------|---------|---------|---------|---------|---------|---------|---------| |4MB Cap| 512 MB | 1024 MB | 1536 MB | 2048 MB | 2560 MB | 3072 MB | 5120 MB | |-------|---------|---------|---------|---------|---------|---------|---------| |user | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | 0.00s | |system | 0.00s | 0.01s | 0.01s | 0.01s | 0.02s | 0.02s | 0.04s | |cpu | 0% | 0% | 0% | 0% | 0% | 0% | 0% | |total | 1.475s | 3.046s | 4.643s | 6.159s | 7.779s | 9.340s | 15.476s | |-------|---------|---------|---------|---------|---------|---------|---------| |copy | 128 | 256 | 384 | 512 | 640 | 768 | 1280 | |commit | 128 | 256 | 384 | 512 | 640 | 768 | 1280 | |-------|---------|---------|---------|---------|---------|---------|---------| If you don't have any objections then I'll put in the 4MB cap with a comment that it's to keep from holding open the RPC slot too long. Then I'll resubmit that along with the xfstests I've written (and one vfs fix). Thoughts? Anna > > --b. > -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
