On Thursday, December 10, 2015 at 3:23:02 PM UTC-5, Riku Bister wrote: > > > >> >> *mount.s3ql --compress none --threads 25 --metadata-upload-interval 3600 >> --allow-other swift://localhost/default /mnt/hubic* >> >> >> i think cpu is ok, did you watch one cpu or total cpu(s) usage? press > number "1" on *top *to open all cores. > > and you might try * --threads 5* or less, i have noticed they do matter, > you must find a best setting what suites you. try maybe first 2 then raise > it. to me was not much difference between 5 or 10, 10 just took alot more > power, but as i cant upload same time so fast then not matter (im using > ownCloud) also using s3ql as samba share to my home network via openvpn > tunnel > Try also different cachesize, lets say 1gig or something.. > *--cachesize=1512000* = 1.5G > > Im using this setting --allow-other --cachesize=1512000 --threads=5 > --metadata-upload-interval=7200 --compres=none --nfs (i have mkfsed this > particular filesystem with 50M *--max-obj-size*) i have 3 filesystems > working same time all have different setting, and this one is mainly for > media. about NFS setting, i have no idea what it does, there not much > documentation about it, so i put it because i use samba, haven't tried yet > without it > > Unfortunately I tried your suggestions and it made no noticeable difference. I can get up to maybe 5MB/sec but it should be about 10MB/sec, because that's what I can get with swiftexplorer.
If I watch the bandwidth used with the program *bmon* I can see it spike over and over, like it's not constantly writing to the mount, but only every few seconds, so it averages out to half the speed it should be. -- You received this message because you are subscribed to the Google Groups "s3ql" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
