AW: Re: WG: AW: AW: AW: WG: Re: read performance much slow er than write

hi Peter!


>cut, but that's all. If you see a difference, then the kernel has a
>problem with its caching algorithm.

After a few tests playing with the option -n on the server side and
different settings on client side I discovered the following, maybe it
helps:

All tests are using 4 enbd-servers with raid5:
- on server side without option -n, on client side without merge_requests:
  read speed about 36MB/s

- on server side with option -n, on client side without merge_requests:
  read speed about 36MB/s

- on server side without option -n, on client side with merge_requests=512:
  read speed about 24MB/s (As meusured already in other tests before)

- on server side with option -n, on client side with merge_requests=512:
  read speed about 46MB/s (As meusured already in other tests before)

It seems that merge_requests slows down read performance significantly if
the option -n on the server is not used. If it is used it speeds up the
performance.

>I imagine a huge cache and a slow processor might be nasty. What was
>the output from "free" like during slow reads? Or meminfo? ANy
>noticable difference wrt fast reads under direct i/o?

free on servers during the test without -n on servers with merge
requests=512 on client with raid5 24,5MB/s read speed:
server1:
             total       used       free     
Mem:       1036524     114496     922028          
-/+ buffers/cache:      30012    1006512
Swap:      1999992          0    1999992

shared    buffers     cached
     0      48616      35868

server2:
             total       used       free     
Mem:       1036524     120364     916160          
-/+ buffers/cache:      32452    1004072
Swap:      1999992          0    1999992

shared    buffers     cached
     0      71496      16416

server3:
             total       used       free     
Mem:       1036524     142476     894048          
-/+ buffers/cache:      34800    1001724
Swap:      1999992          0    1999992

shared    buffers     cached
     0      90800      16876

server4:
             total       used       free     
Mem:       1036524     161992     874532          
-/+ buffers/cache:      36644     999880
Swap:      1999992          0    1999992

shared    buffers     cached
     0     113952      11396


>You can compare speed with/without direct i/o on the server side using a
>simple c program. I'll write it if you like.

Yes, please since I am not so firm with c.

>How can you use "2 clients"?

Oh, sorry I was not clear, I meant 2 different server machines (computers)
running enbd-clients both running with the same setup but using different
partions on the servers.

>You need to measure much more at the same time for the data to be more
>meaningful. In particular network and cpu loading values are needed.

I try to do all tests in the middle of next week because I will be away till
wednesday.

>Maybe Arne can tell you how to make such measurements.

Could you please connect me to him so I can ask him to help me out with
those tests?


Greetings,

Christian