On the subject of which statistics to display on a cifs/smb3 mount - one other example for comparison, mountstats. Note that nfs is the only filesystem which seems to display detailed information in /proc//mountstats See below sample output e.g. "cat /proc/1/mountstats" (would this be useful to add for cifs.ko? ie implementing "show_stats" superblock method as nfs does to display more detailed smb3 specific protocol statistics) device vmware-vmblock mounted on /run/vmblock-fuse with fstype fuse.vmware-vmblock device //localhost/test mounted on /mnt with fstype cifs device localhost:/nfsexport mounted on /mnt1 with fstype nfs4 statvers=1.1 opts: rw,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,acregmin=3,acregmax=60,acdirmin=30,acdirmax=60,hard,proto=tcp,timeo=600,retrans=2,sec=sys,clientaddr=127.0.0.1,local_lock=none age: 1191 impl_id: name='',domain='',date='0,0' caps: caps=0x1fbffff,wtmult=512,dtsize=32768,bsize=0,namlen=255 nfsv4: bm0=0xfdffbfff,bm1=0x40f9be3e,bm2=0x20803,acl=0x3,sessions,pnfs=not configured sec: flavor=1,pseudoflavor=1 events: 11 9 0 0 6 0 19 1 0 19 0 3 12 0 3 2 0 2 0 0 1 0 0 0 0 0 0 bytes: 471859200 5 0 0 503316480 5 122880 1 RPC iostats version: 1.0 p/v: 100003/4 (nfs) xprt: tcp 949 0 1 0 56 541 541 0 3932 0 11 13 3391 per-op statistics NULL: 0 0 0 0 0 0 0 0 READ: 480 480 0 115200 503366400 1216 1606 2827 WRITE: 1 1 0 268 176 0 4 4 COMMIT: 0 0 0 0 0 0 0 0 OPEN: 1 1 0 340 372 0 6 6 OPEN_CONFIRM: 0 0 0 0 0 0 0 0 OPEN_NOATTR: 1 1 0 292 352 0 0 0 OPEN_DOWNGRADE: 0 0 0 0 0 0 0 0 CLOSE: 2 2 0 480 288 0 0 0 SETATTR: 0 0 0 0 0 0 0 0 FSINFO: 1 1 0 188 164 0 0 0 ... On Tue, Jul 31, 2018 at 4:17 PM Steve French wrote: > > > On Tue, Jul 31, 2018 at 3:42 PM Volker Lendecke > wrote: > >> On Tue, Jul 31, 2018 at 03:21:02PM -0500, Steve French via >> samba-technical wrote: >> >> > There are other statistics that can be dumped in Windows redirector, but >> > those seem to be the most visible (those available via "net statistics >> > workstation"). Any other suggested statistics? >> >> Response times. I know they are more difficult to track, but a little >> logarithmic histogram might be doable. >> >> > Probably doable. What do you think of the output of iostat (here is > default iostat output)? > > $ iostat /dev/nvme0n1 > Linux 4.18.0-041800rc7-generic (smf-Thinkpad-P51) 07/31/2018 > _x86_64_ (8 CPU) > > avg-cpu: %user %nice %system %iowait %steal %idle > 11.74 0.19 5.31 0.05 0.00 82.71 > > Device tps kB_read/s kB_wrtn/s kB_read kB_wrtn > nvme0n1 5.69 59.45 45.30 4187172 3190324 > > > Presumably nfsstats are not any more helpful - here is sample client stats > output from nfsstat > > $ nfsstat -c > Client rpc stats: > calls retrans authrefrsh > 522 0 522 > > Client nfs v4: > null read write commit > open > 1 0% 480 91% 1 0% 0 0% > 1 0% > open_conf open_noat open_dgrd close > setattr > 0 0% 1 0% 0 0% 2 0% > 0 0% > fsinfo renew setclntid confirm > lock > 3 0% 0 0% 0 0% 0 0% > 0 0% > lockt locku access getattr > lookup > 0 0% 0 0% 2 0% 12 2% > 2 0% > lookup_root remove rename link > symlink > 1 0% 0 0% 0 0% 0 0% > 0 0% > create pathconf statfs readlink > readdir > 0 0% 2 0% 2 0% 0 0% > 2 0% > server_caps delegreturn getacl setacl > fs_locations > 5 0% 0 0% 0 0% 0 0% > 0 0% > rel_lkowner secinfo fsid_present exchange_id > create_session > 0 0% 0 0% 0 0% 2 0% > 1 0% > destroy_session sequence get_lease_time reclaim_comp > layoutget > 0 0% 0 0% 0 0% 1 0% > 0 0% > getdevinfo layoutcommit layoutreturn secinfo_no > test_stateid > 0 0% 0 0% 0 0% 1 0% > 0 0% > free_stateid getdevicelist bind_conn_to_ses destroy_clientid > seek > 0 0% 0 0% 0 0% 0 0% > 0 0% > allocate deallocate layoutstats clone > 0 0% 0 0% 0 0% 0 0% > > -- > Thanks, > > Steve > -- Thanks, Steve