Performance

→ please see performance for HowTos!

The test was done with two fast PCs wit GBit-Nic and the 1043ND loaded with 10.03 between them. Netperf was used to measure the troughput with TCP and TCP_MAERTS tests.

• Bridged: 100 MByte/s
• Routed: 23 Mbyte/s
• Routed without Firewall and Conntrack: 29MB/s

Another test using two PCs with GBit-Nic and 1043ND v2.1 loaded with 15.05 in between. Using CrystalDiskMark and NTFS fileshare (Win7):

LAN↔LAN:

• READ: 117,6 MB/s
• WRITE: 117,6 MS/s

yep, we sure are are running 1Gbit, and the PCs will not be the bottleneck

LAN↔WAN (default firewall and NAT settings):

• Upload: 55.89 MB/s
• Download: 25.74 MB/s

USB performance is measured with unformatted partition on HDD connected via SATA-USB controller Prolific PL3507. Benchmark tool is dd running on the router and using /dev/zero for read test and /dev/null for write test.

• Read: 22.6 MByte/s
• Write: 24.4 MByte/s

Remember, it is “in vitro” conditions results, you will NOT get such performance via NFS/Samba/FTP server and with real file system.

XChesser has posted on the forum his comparative test of performance for different file systems under Backfire 10.03.1-RC4.

Testing methodology from the forum post: the objective was to measure the speed of filesystems on the router with OpenWrt excluding the impact of other factors as much as possible. My TP-Link WR1043ND runs OpenWrt Backfire 10.03.1-RC4. Kernel version: 2.6.32.25-1. As an external USB-drive I used Samsung HD204UI connected to the router via SATA-USB controller based on Prolific Technology PL3507 chip. About 15 partitions with different FS and 1GB in size were created with the default settings. For comparison I also made a read/write test with unformatted partition - this rate can be considered as the limit of the hardware, because not any FS-driver is involved in this case.

As benchmark tool full version of dd was used. Thus the resources of the router were not spent on data transmission over the network and the corresponding server. To minimize source file read/write expenses on the router side, /dev/zero and /dev/null weer used, respectively. Multiple tests with a file of 800 MB on Ext2 gave 3% spread only so this file size was used in all tests subsequently. All filesystems were tested with 2 values of the read/write block size: 4 KB and 2 MB. For the purity of the experiment unnecessary services and servers were unloaded for benchmark time; no router accesses except for SSH took place. In this state, the average CPU Idle was 99.7%.

0_1d0abc_a2e837d9_orig

These figures come from a TP-Link WR1043ND v1.8 running backfire 10.03.1-RC4 and samba 3.0.4. The HDD is an NTFS formatted 7200 RPM 3.5“ disk which is capable of read performance in excess of 30 MB/s when plugged into a computer. A large (several hundred MiB) file is copied from and to the router, the total time is recorded with a stopwatch; the average transfer rate is eventually computed:

• Read: 8.2 MiB/s (stock firmware reaches about 6.5 MiB/s under similar conditions)
• Write: 3.1 MiB/s

Bear in mind ntfs-3g is a CPU intensive file system, that doesn't match the performance of the Windows driver even with reasonably powerful hardware (e.g. dual core AMD64 Linux system). It also known to be very sensitive to fragmentation.

Transfer rate from USB2 stick to computer via 1000 Mbit/s ethernet connection using vfat and NFS was of 7.7 MBytes/s.