> > If I do a stress test of i/o to the san - > an 8G cp from the san to the internal drive and > visa versa - I see halts of simple > montoring programs . > First thing I would do is to try to figure out if this "halt" is due the disks on your RHEL system being very busy and not being able to respond in a timely manner to you user requested programs. I'd also want to know if the "halt" is affecting the i/o from your system to the SAN as well. With this info you should be able to get a better idea of where the bottleneck is and be able to know what you're supposed to try to tune. To test if the halt is due to the disks in your RHEL box you can use some dd command like below, this way you're not stress testing your disks, you're just stress testing the fiber connection and the san. # Write test dd if=/dev/zero of=/mnt/mountpoint/file bs=4096 count=1572864 # Read speed test dd if=/mnt/mountpoint/file of=/dev/null bs=4096 To see if the "halt" is also stopping i/o you should be able to get your SAN admin or you (if the SAN admin is you) to look at the i/o on the said san volume and watch it during the stress testing, to see if it drops during the "halts". Good luck and we'd be interested to hear what you eventually find, Romeo Theriault System Administrator Information Technology Services -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
