On Fri, 16 Jun 2017 14:21:34 +0200 Fabian Grünbichler <f.gruenbich...@proxmox.com> wrote:
> > one big point that I haven't really looked at is whether you need to > guard some operations with locks. especially the calls to iscsiadm are > triggered in a lot of code paths, and I don't know whether those are > problematic when executed concurrently or not. some of the higher level I have not found any evidence that locks should be needed since, by nature, the scsi subsystem is thread safe. This is not a problem I have seen so far. > > also haven't checked for how it would handle an already existing local > (open-)iscsi setup.. did you? > My proof of concept (POC) was tested on an existing iscsi setup added libiscsi to the pool. There were no interfering with existing setup what so ever. The clever thing in my setup is that all interaction with the scsi subsystem is closely tied to a specific session which is invisible to other sessions. You can compared it to a database transaction. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael <at> rasmussen <dot> cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir <at> datanom <dot> net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir <at> miras <dot> org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Come quickly, I am tasting stars! -- Dom Perignon, upon discovering champagne.
pgplms2CWxo9u.pgp
Description: OpenPGP digital signature
_______________________________________________ pve-devel mailing list pve-devel@pve.proxmox.com https://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
