nuaa_liuben
2012-Oct-11 07:03 UTC
[Gluster-users] samba performance downgrade with glusterfs backend
Hi folks, We found that samba performance downgrade a lot with glusterfs backend. volume info as followed, Volume Name: vol1 Type: Distribute Status: Started Number of Bricks: 1 Transport-type: tcp Bricks: Brick1: pana53:/data/ Options Reconfigured: auth.allow: 192.168.* features.quota: on nfs.disable: on Use dd (bs=1MB) or iozone (block=1MB) to test write performance, about 400MB/s. #dd if=/dev/zero of=dd.dat bs=1MB count=1k 1024+0 records in 1024+0 records out 1024000000 bytes (1.0 GB) copied, 2.6142 s, 392 MB/s But exporting with samba, use 4 Win7 clients to test with SANergy/Iometer, write performance only about 170MB/s. Command line used: iozone -s 1g -r 1m -i 0 -t 4 Output is in Kbytes/sec Time Resolution = 0.000001 seconds. Processor cache size set to 1024 Kbytes. Processor cache line size set to 32 bytes. File stride size set to 17 * record size. Throughput test with 4 processes Each process writes a 1048576 Kbyte file in 1024 Kbyte records Children see throughput for 4 initial writers = 171295.03 KB/sec Parent sees throughput for 4 initial writers = 170732.54 KB/sec Min throughput per process = 42684.70 KB/sec Max throughput per process = 42991.59 KB/sec Avg throughput per process = 42823.76 KB/sec Min xfer = 1041408.00 KB #smb.conf [global] log file = /var/log/samba/log.%m max log size = 50 guest account = nobody map to guest = bad password posix locking = no max protocol = SMB2 socket options = TCP_NODELAY IPTOS_LOWDELAY [test] comment = None path = /mnt/glusterfs browseable = yes read only = no hosts allow = hosts deny = guest ok = yes Anyone meet such issues? thanks a lot nuaa_liuben -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20121011/cdcf9328/attachment.html>