MoshiachNow
IS-IT--Management
HI,
Lately I have dealt with slow disk write on 6E4 (p630) and 6E3.
I'm talking about 73 GB IBM disks that come standard with these machines.
Reading from them was fine - above 50MB/sec.However writing was maximum 9MB/sec.
At the beginning I have connected an external tower with Seagate disks to these boxes,and it gave me 50 MB/sec write,so I thought I'd have to split the bus between the system and data disks.
But looking deeper into the issue together with an engineering guide - we have found that the IBM "slow" disks start to "fly" (50-60 MB/s) once we enable the write cache by playing with the IBM stanza !
It appears that the IBM disks are coming with the write cache closed - both on the disk and as part of the operating system (stanza).
(We still did not finalized the good stanza for IBM disks - the above good results were just the results of playing with the numbers in the stanza that affect the disk write.I hope to publish the final stanza once we have it).
"Long live king Moshiach !"
Lately I have dealt with slow disk write on 6E4 (p630) and 6E3.
I'm talking about 73 GB IBM disks that come standard with these machines.
Reading from them was fine - above 50MB/sec.However writing was maximum 9MB/sec.
At the beginning I have connected an external tower with Seagate disks to these boxes,and it gave me 50 MB/sec write,so I thought I'd have to split the bus between the system and data disks.
But looking deeper into the issue together with an engineering guide - we have found that the IBM "slow" disks start to "fly" (50-60 MB/s) once we enable the write cache by playing with the IBM stanza !
It appears that the IBM disks are coming with the write cache closed - both on the disk and as part of the operating system (stanza).
(We still did not finalized the good stanza for IBM disks - the above good results were just the results of playing with the numbers in the stanza that affect the disk write.I hope to publish the final stanza once we have it).
"Long live king Moshiach !"
