Hi Suresh,

 

has your server installed a HP SmartArray P400 array controller? This controller has 256 or (what I believe) 512 MB battery buffered cache. Testing with a 100 MB file will result in Testing SA P400 cache response time. Testing with a 1 GB file will still result in 50 % SA P400 cache response time.

 

My experience is that a FC connected Filer will not reach throughput of a SA P400 cache while testing with small file sizes. But you may try your tests with much more data. Let’s say 10 processes in parallel with 1 GB each. You may also write more data to your filer than your filer has memory installed (= 16 GB) to see it’s memory to disk write speed.

 

But for all this the two questions are: What do you want to benchmark?  and: Is this reflecting your workload?

 

Best Regards

 

 

 

Here are some numbers from crystaldiskmark.  On crystaldiskmark the max file size is 1000Mb.  The host has 64G of RAM and 8 six-core processors.

 

1.       DL785G6, 100Mb file local disk

a.       Seq                        Read      216.5MB/sec     Write  78.4MB/sec

b.      Random 512k     Read      58.14MB/sec      Write 301.9MB/sec

c.       Random 4k         Read      26.2MB/sec        Write 41.3MB/sec

2.       DL785G6 100MB file on Filer LUN (NTFS)

a.       Seq                        Read      175.1MB/sec      Write 100.7MB/sec

b.      Random 512k     Read      103.7MB/sec      Write    71.44MB/sec

c.       Random 4k         Read      15.7MB/sec        Write    7.6MB/sec

3.       DL785G6 1000MB file on local disk

a.       Seq                        Read      236.8MB/sec      Write 92.7MB/sec

b.      Random 512k     Read      49.71MB/sec      Write 217.2MB/sec

c.       Random 4k         Read      1.33MB/sec        Write  20.63

4.       DL785G6 1000Mb file on filer LUN (NTFS)

a.       Seq                        Read      164MB/sec      Write 98.8MB/sec

b.      Random 512k     Read      101.1MB/sec      Write    63.2MB/sec

c.       Random 4k         Read      13.9MB/sec        Write    7.8MB/sec

 

Suresh

 

 

 

Suresh,

 

Performance benchmarking is a science that involves many variables.  I am not familiar with CrystalDiskMark but I just downloaded the source for 3.0 RC2 and will have a look to see how applicable it could be to a filer vs local disk comparison.  Can you add some more details about your configuration?  (any options you run with the test, specs/model of the server including controller/RAID card(s), OS on the server, disk model in the server, disks in the filer, model of the filer, ONTAP rev, etc).  A lot of detail is going to be required to make any headway or recommendations for a valid test.

 

Thank you,

Tim

 

 

I’m using the free tool Crystaldiskmark to do some  I/O comparison between local disk and our filers. On at least one system  (SAN connected), the local disk (6 disks in RAID1) consistently comes out ahead in both read and write. Filer is lightly loaded, and this is on a 56 disk aggregate.   I’m kind of stumped on this one, and would like to know if:

 

a)      Are there any other commonly used benchmarks which I can try with the filers? 

b)      This is on a 2G FC SAN.  How much improvement can I expect with 4G or 8G?

 

Thanks

Suresh