Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Mon, 19 Jun 2017 20:54:18 +0200 Michael Rasmussen wrote: > Found it ;-) No need to change any code in LXC.pm > Cause: The code I added to circumvent the FreeNAS bug was doing a > rescan of the session which deactivated the volume before the command > unshare -m -- sh -c

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Mon, 19 Jun 2017 12:57:24 +0200 Fabian Grünbichler wrote: > > yes, that or debugging what mount is actually doing seems like the way > to go. if it fails as well with "-t nonfs", then I think you need to > check your plugin code for further timing problems. >

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Mon, 19 Jun 2017 13:06:48 +0200 "datanom.net" wrote: > > > I will try to add an udevadm trigger/settle before returning from > > > activate_volume in the code I see whether this solves the problem. > Forgot I already do this: +$os_request->(['iscsiadm', '-m',

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On 2017-06-19 12:57, Fabian Grünbichler wrote: but it seems like there is some kind of test run earlier, because when I manually symlink an actual block device to a path with a colon in it, mounting it works as expected, both with run_command and manually. maybe the root cause is still

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Mon, Jun 19, 2017 at 11:59:10AM +0200, datanom.net wrote: > On 2017-06-19 09:33, Fabian Grünbichler wrote: > > > > AFAICT, mount will always guess NFS if there is a ':' anywhere in the > > source path: > > > > No, mount only guess NFS if run through PVE::Tools:run_command. Running the >

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On 2017-06-19 09:33, Fabian Grünbichler wrote: AFAICT, mount will always guess NFS if there is a ':' anywhere in the source path: No, mount only guess NFS if run through PVE::Tools:run_command. Running the exact same command as produced in PVE::LXC::API2 from the command line works as

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On 2017-06-19 09:35, Fabian Grünbichler wrote: I was afraid of something like this.. the whole API does not seem very though through unfortunately. let's see whether the paging is doable, and otherwise document this potential pitfall (e.g., having more than 1 snapshots on a busy big storage

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On 2017-06-19 09:47, Fabian Grünbichler wrote: I was thinking more about logical races, along the lines of the following: - no existing session - activate volume A A- get_active_luns (note: this will already create an empty session) A- freenas_get_lun_number - concurrent activate volume B B-

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

ah, now I get it. please correct me if I misunderstood something (again): - get all targettoextent mappings and loop over them - only look at those with the desired target and with a lunid that is in the "special range" reserved for snapshots and state volumes - if the volume has an extent,

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Fri, Jun 16, 2017 at 03:42:07PM +0200, datanom.net wrote: > Hi Fabian, > > Most of your input I will study this weekend but a can give a quick answers > to a few comments which follows below. > > On 2017-06-16 14:21, Fabian Grünbichler wrote: > > > next unless $zvol->{name} =~

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Fri, Jun 16, 2017 at 10:28:39PM +0200, Michael Rasmussen wrote: > On Fri, 16 Jun 2017 14:21:34 +0200 > Fabian Grünbichler wrote: > > > > > one big point that I haven't really looked at is whether you need to > > guard some operations with locks. especially the

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Fri, Jun 16, 2017 at 10:50:57PM +0200, Michael Rasmussen wrote: > On Fri, 16 Jun 2017 14:21:34 +0200 > Fabian Grünbichler wrote: > > > > my $limit = 1; # limit for get requests > > > > this seems kind of arbitrary? if you really need such limit, you also > >

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Sun, Jun 18, 2017 at 12:47:30AM +0200, Michael Rasmussen wrote: > Hi Fabian, > > On Fri, 16 Jun 2017 14:21:34 +0200 > Fabian Grünbichler wrote: > > > > sub volume_resize { > > > my ($class, $scfg, $storeid, $volname, $size, $running) = @_; > > > > > > my

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

Hi Fabian, On Fri, 16 Jun 2017 14:21:34 +0200 Fabian Grünbichler wrote: > > sub volume_resize { > > my ($class, $scfg, $storeid, $volname, $size, $running) = @_; > > > > my ($vtype, $name, $vmid) = $class->parse_volname($volname); > > > > my $run =

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Sat, Jun 17, 2017 at 5:05 PM, Michael Rasmussen wrote: > On Sat, 17 Jun 2017 13:33:38 +0200 > Andreas Steinel wrote: > >> >> You said that you do not have the required hardware, however, you do >> not need additional hardware for that. >> >> I can

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Sat, 17 Jun 2017 13:33:38 +0200 Andreas Steinel wrote: > > You said that you do not have the required hardware, however, you do > not need additional hardware for that. > > I can provide you with a multipath configuration if you like. iSCSI is > not different to

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Sat, Jun 17, 2017 at 10:33 AM, Michael Rasmussen wrote: > On Sat, 17 Jun 2017 10:15:25 +0200 > Andreas Steinel wrote: > >> Mi Michael, >> >> On Fri, Jun 16, 2017 at 3:10 AM, Michael Rasmussen wrote: >> > MPIO is not tested since I

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Sat, 17 Jun 2017 10:15:25 +0200 Andreas Steinel wrote: > Mi Michael, > > On Fri, Jun 16, 2017 at 3:10 AM, Michael Rasmussen wrote: > > MPIO is not tested since I don't have the required hardware to do so. > > Is there more than adding another IP to

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

Mi Michael, On Fri, Jun 16, 2017 at 3:10 AM, Michael Rasmussen wrote: > MPIO is not tested since I don't have the required hardware to do so. Is there more than adding another IP to the FreeNAS machine? I've done that on a Debian-based target and it automatically logged into

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Fri, 16 Jun 2017 14:21:34 +0200 Fabian Grünbichler wrote: > > my $limit = 1; # limit for get requests > > this seems kind of arbitrary? if you really need such limit, you also > need to implement pagination or something like it, otherwise you are in >

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Fri, 16 Jun 2017 14:21:34 +0200 Fabian Grünbichler 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

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

Hi Fabian, Most of your input I will study this weekend but a can give a quick answers to a few comments which follows below. On 2017-06-16 14:21, Fabian Grünbichler wrote: next unless $zvol->{name} =~ /^(base|vm)-(\d+)-disk-\d+$/; $vmid = $2; $parent = undef;

Re: [pve-devel] [PATCH pve-storage 0/5] v5 fix regression and indentation

On Fri, Jun 16, 2017 at 03:10:27AM +0200, Michael Rasmussen wrote: > v5 regression and indentation fix > V4 Adds support for creating snapshot backups of LXC containers. This more or > less makes the plugin feature complete. Only outstanding feature depends on a > bug fix in the FreeNAS API