Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi, > I get the feeling that someone needs to read up about caching, > especially Linux page cache, Linux IO stack, > and where DRBD is located in there. never hurts. > We typically have > [ applications ] > [ application and library buffers ] > [ file systems ] > [ page cache ] > [ block layer ] <=== DRBD lives here > `--- drbd replication via TCP ---> remote "DISK". > [ "DISK" ] I see. So my guess that the controller cache is to blame was probably off the mark, no? More-than-NIC-capacity throughput could only be facilitated by write cacheing going on in the page cache or above, as per your diagram. But I guess I'm still sort of pole-poking the issue. Thanks for clearing things up! Felix