This site is currently read-only as we are migrating to Oracle Forums for an improved community experience. You will not be able to initiate activity until January 31st, when you will be able to use this site as normal.

    Forum Stats

  • 3,890,489 Users
  • 2,269,776 Discussions
  • 7,916,824 Comments

Discussions

How to run VDBench filesystem curveIO with multiple xfersize using a single parmfile?

User_PSVHB
User_PSVHB Member Posts: 1 Green Ribbon

parm-1 is for filesystem sequential write & read workloads and it works just fine.

parm-2 replacing fwdrate=max to fwdrate=curve in rd didn't work for filesystem curveio.

# In past experiments/attempts to make filesystem curveio work, we had reached scenario where post initial/first fwdrate=curve i.e. 'max IOPS' run, the rest of curveIO % were doing only 4k random reads whatever be the intended workload and xfersize. 

# It took us a while to make necessary changes (i.e. not using -d128, using xfersize= and operation= in fwd= in place of rd=) to make filesystem curve IO work with desired xfersize and operation as intended. 

# Problem Statement: But now using single parmfile, curveIO works for single(1) xfersize parameter. 

Question/Request: 

How can we specify the filesystem curveIO to be run with multiple xfersize using a single parmfile? (e.g. using a single parmfile, we want to run curveIO for 256k, 512k, and 1024k IO block sizes.)