> > > Hello, > > > > Let's assume that we have a resource which contains several volumes. > > When an on-line verification is launched, all volumes are verified > > together, but it seems that the job is made by only one thread. > > The bottleneck in the on-line verification are not the disks, but the > > CPU, as it must deal with many (for example) crc32c operations. > > > > So my question is, could we think about an implementation which would > > start one verification thread for each volume ? > > It would dramatically speed-up on-line verification (for my own case, > > with 4 volumes in 1 resource, it would decrease verification time from > > 56 hours to 14 hours). > > iirc, when we implemented it, the kernel crypto api was supposed > to get asynchronous hash support "soon", so we did not bother to add our > own threading/async layer over this. > > Guess that "soon" is very relative. > > Yes, we can add that as feature request > somewhere on our (not so short) todo list. >
Lars, Thank you very much for your explanation and for taking this request into account ! Best regards, Ben
_______________________________________________ drbd-user mailing list [email protected] http://lists.linbit.com/mailman/listinfo/drbd-user
