dlandon Posted March 12, 2016 Share Posted March 12, 2016 I must be missing something here. My test server won't auto start with no parity drives assigned. Link to comment
dlandon Posted March 12, 2016 Share Posted March 12, 2016 Getting the following error on the console when shutting down the server: /etc/rc.d/rc.6: line 262: [: missing ']' Link to comment
Furby8704 Posted March 12, 2016 Share Posted March 12, 2016 i squealed like a little girl hahaha time for a text build Link to comment
Squid Posted March 12, 2016 Share Posted March 12, 2016 Where can we find the documentation on the new xml attributes for docker containers (ie: hidden, required, etc etc.) Or, do we just have to play around with the options in dockerMan and see what user templates it creates with the various options? Link to comment
dlandon Posted March 12, 2016 Share Posted March 12, 2016 NFS appears to be failing. Log entries: Mar 11 23:49:08 Tower rpc.statd[8155]: Version 1.3.3 starting Mar 11 23:49:08 Tower sm-notify[8156]: Version 1.3.3 starting Mar 11 23:49:08 Tower sm-notify[8156]: Already notifying clients; Exiting! Mar 11 23:49:08 Tower rpc.statd[8155]: failed to create RPC listeners, exiting Mar 11 23:49:08 Tower root: Starting NFS server daemons: Mar 11 23:49:08 Tower root: /usr/sbin/exportfs -r Mar 11 23:49:08 Tower root: /usr/sbin/rpc.nfsd 8 Mar 11 23:49:08 Tower root: rpc.nfsd: address family AF_INET6 not supported by protocol TCP Mar 11 23:49:08 Tower root: rpc.nfsd: unable to set any sockets for nfsd Mar 11 23:49:08 Tower root: /usr/sbin/rpc.mountd Mar 11 23:49:08 Tower root: rpc.mountd: svc_tli_create: could not open connection for udp6 Mar 11 23:49:08 Tower root: rpc.mountd: svc_tli_create: could not open connection for tcp6 Mar 11 23:49:08 Tower root: rpc.mountd: svc_tli_create: could not open connection for udp6 Mar 11 23:49:08 Tower root: rpc.mountd: svc_tli_create: could not open connection for tcp6 Mar 11 23:49:08 Tower root: rpc.mountd: svc_tli_create: could not open connection for udp6 Mar 11 23:49:08 Tower root: rpc.mountd: svc_tli_create: could not open connection for tcp6 Mar 11 23:49:08 Tower rpc.mountd[8166]: mountd: No V2 or V3 listeners created! Link to comment
PeterB Posted March 12, 2016 Share Posted March 12, 2016 Mar 11 23:49:08 Tower rpc.mountd[8166]: mountd: No V2 or V3 listeners created! No nfs v4 support built in yet? Link to comment
danioj Posted March 12, 2016 Share Posted March 12, 2016 On my test server, from the minute I turn it on I get a recycling set of messages for each device in the syslog over and over again like this: Mar 12 18:21:02 test kernel: sdb: sdb1 Mar 12 18:21:02 test emhttp: shcmd (1141): rmmod md-mod |& logger Mar 12 18:21:02 test kernel: md: unRAID driver removed Mar 12 18:21:02 test emhttp: shcmd (1142): modprobe md-mod super=/boot/config/super.dat |& logger Mar 12 18:21:02 test kernel: md: unRAID driver 2.6.0 installed Mar 12 18:21:02 test emhttp: Pro key detected, GUID: 0718-048A-07B2-1607419A1FD1 FILE: /boot/config/Pro.key Mar 12 18:21:02 test emhttp: Device inventory: Mar 12 18:21:02 test emhttp: shcmd (1143): udevadm settle Mar 12 18:21:02 test emhttp: TDKMedia_Gold_Flash_07B21607419A1FD1-0:0 (sda) 7811040 Mar 12 18:21:02 test emhttp: WDC_WD800AAJS-00PSA0_WD-WMAPA1162808 (sdb) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800JD-60LSA5_WD-WMAM9MF49542 (sdc) 78150712 Mar 12 18:21:02 test emhttp: ST380815AS_5RW18GXH (sdd) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800JD-60LSA5_WD-WMAM9PV29584 (sde) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800AAJS-00PSA0_WD-WMAPA1147612 (sdf) 78150712 Mar 12 18:21:02 test kernel: mdcmd (1): import 0 Mar 12 18:21:02 test kernel: md: import_slot: 0 empty Mar 12 18:21:02 test kernel: mdcmd (2): import 1 Mar 12 18:21:02 test kernel: md: import_slot: 1 missing Mar 12 18:21:02 test kernel: mdcmd (3): import 2 Mar 12 18:21:02 test kernel: mdcmd (4): import 3 Mar 12 18:21:02 test kernel: mdcmd (5): import 4 Mar 12 18:21:02 test kernel: mdcmd (6): import 5 Mar 12 18:21:02 test kernel: mdcmd (7): import 6 Mar 12 18:21:02 test kernel: mdcmd (: import 7 Mar 12 18:21:02 test kernel: mdcmd (9): import 8 Mar 12 18:21:02 test kernel: mdcmd (10): import 9 Mar 12 18:21:02 test kernel: mdcmd (11): import 10 Mar 12 18:21:02 test kernel: mdcmd (12): import 11 Mar 12 18:21:02 test kernel: mdcmd (13): import 12 Mar 12 18:21:02 test emhttp: import 30 cache device: no device Mar 12 18:21:02 test emhttp: import flash device: sda Mar 12 18:21:02 test kernel: mdcmd (14): import 13 Mar 12 18:21:02 test kernel: mdcmd (15): import 14 Mar 12 18:21:02 test kernel: mdcmd (16): import 15 Mar 12 18:21:02 test kernel: mdcmd (17): import 16 Mar 12 18:21:02 test kernel: mdcmd (18): import 17 Mar 12 18:21:02 test kernel: mdcmd (19): import 18 Mar 12 18:21:02 test kernel: mdcmd (20): import 19 Mar 12 18:21:02 test kernel: mdcmd (21): import 20 Mar 12 18:21:02 test kernel: mdcmd (22): import 21 Mar 12 18:21:02 test kernel: mdcmd (23): import 22 Mar 12 18:21:02 test kernel: mdcmd (24): import 23 Mar 12 18:21:02 test kernel: mdcmd (25): import 24 Mar 12 18:21:02 test kernel: mdcmd (26): import 25 Mar 12 18:21:02 test kernel: mdcmd (27): import 26 Mar 12 18:21:02 test kernel: mdcmd (28): import 27 Mar 12 18:21:02 test kernel: mdcmd (29): import 28 Mar 12 18:21:02 test kernel: mdcmd (30): import 29 Mar 12 18:21:02 test kernel: md: import_slot: 29 empty Mar 12 18:21:02 test emhttp: shcmd (1149): /etc/rc.d/rc.nfsd start |& logger Mar 12 18:22:37 test login[6398]: ROOT LOGIN on '/dev/tty1' Mar 12 18:23:33 test in.telnetd[29155]: connect from 192.168.1.2 (192.168.1.2) Mar 12 18:23:35 test login[29156]: ROOT LOGIN on '/dev/pts/1' from 'main.danioj.lan' To indicate what I am doing, go here: https://lime-technology.com/forum/index.php?topic=47427.0 Currently pre clearing drives using gfjardim's new preclear plugin / script: http://lime-technology.com/forum/index.php?topic=39985.msg453938#msg453938 BUT from the syslog it doesn't appear to the cause IMHO because what is in the log seems to start as soon as the server starts up. No disks assigned at the current time and as such Array not started. test-diagnostics-20160312-1830.zip Link to comment
itimpi Posted March 12, 2016 Share Posted March 12, 2016 On my test server, from the minute I turn it on I get a recycling set of messages for each device in the syslog over and over again like this: Mar 12 18:21:02 test kernel: sdb: sdb1 Mar 12 18:21:02 test emhttp: shcmd (1141): rmmod md-mod |& logger Mar 12 18:21:02 test kernel: md: unRAID driver removed Mar 12 18:21:02 test emhttp: shcmd (1142): modprobe md-mod super=/boot/config/super.dat |& logger Mar 12 18:21:02 test kernel: md: unRAID driver 2.6.0 installed Mar 12 18:21:02 test emhttp: Pro key detected, GUID: 0718-048A-07B2-1607419A1FD1 FILE: /boot/config/Pro.key Mar 12 18:21:02 test emhttp: Device inventory: Mar 12 18:21:02 test emhttp: shcmd (1143): udevadm settle Mar 12 18:21:02 test emhttp: TDKMedia_Gold_Flash_07B21607419A1FD1-0:0 (sda) 7811040 Mar 12 18:21:02 test emhttp: WDC_WD800AAJS-00PSA0_WD-WMAPA1162808 (sdb) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800JD-60LSA5_WD-WMAM9MF49542 (sdc) 78150712 Mar 12 18:21:02 test emhttp: ST380815AS_5RW18GXH (sdd) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800JD-60LSA5_WD-WMAM9PV29584 (sde) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800AAJS-00PSA0_WD-WMAPA1147612 (sdf) 78150712 Mar 12 18:21:02 test kernel: mdcmd (1): import 0 Mar 12 18:21:02 test kernel: md: import_slot: 0 empty Mar 12 18:21:02 test kernel: mdcmd (2): import 1 Mar 12 18:21:02 test kernel: md: import_slot: 1 missing Mar 12 18:21:02 test kernel: mdcmd (3): import 2 Mar 12 18:21:02 test kernel: mdcmd (4): import 3 Mar 12 18:21:02 test kernel: mdcmd (5): import 4 Mar 12 18:21:02 test kernel: mdcmd (6): import 5 Mar 12 18:21:02 test kernel: mdcmd (7): import 6 Mar 12 18:21:02 test kernel: mdcmd (: import 7 Mar 12 18:21:02 test kernel: mdcmd (9): import 8 Mar 12 18:21:02 test kernel: mdcmd (10): import 9 Mar 12 18:21:02 test kernel: mdcmd (11): import 10 Mar 12 18:21:02 test kernel: mdcmd (12): import 11 Mar 12 18:21:02 test kernel: mdcmd (13): import 12 Mar 12 18:21:02 test emhttp: import 30 cache device: no device Mar 12 18:21:02 test emhttp: import flash device: sda Mar 12 18:21:02 test kernel: mdcmd (14): import 13 Mar 12 18:21:02 test kernel: mdcmd (15): import 14 Mar 12 18:21:02 test kernel: mdcmd (16): import 15 Mar 12 18:21:02 test kernel: mdcmd (17): import 16 Mar 12 18:21:02 test kernel: mdcmd (18): import 17 Mar 12 18:21:02 test kernel: mdcmd (19): import 18 Mar 12 18:21:02 test kernel: mdcmd (20): import 19 Mar 12 18:21:02 test kernel: mdcmd (21): import 20 Mar 12 18:21:02 test kernel: mdcmd (22): import 21 Mar 12 18:21:02 test kernel: mdcmd (23): import 22 Mar 12 18:21:02 test kernel: mdcmd (24): import 23 Mar 12 18:21:02 test kernel: mdcmd (25): import 24 Mar 12 18:21:02 test kernel: mdcmd (26): import 25 Mar 12 18:21:02 test kernel: mdcmd (27): import 26 Mar 12 18:21:02 test kernel: mdcmd (28): import 27 Mar 12 18:21:02 test kernel: mdcmd (29): import 28 Mar 12 18:21:02 test kernel: mdcmd (30): import 29 Mar 12 18:21:02 test kernel: md: import_slot: 29 empty Mar 12 18:21:02 test emhttp: shcmd (1149): /etc/rc.d/rc.nfsd start |& logger Mar 12 18:22:37 test login[6398]: ROOT LOGIN on '/dev/tty1' Mar 12 18:23:33 test in.telnetd[29155]: connect from 192.168.1.2 (192.168.1.2) Mar 12 18:23:35 test login[29156]: ROOT LOGIN on '/dev/pts/1' from 'main.danioj.lan' To indicate what I am doing, go here: https://lime-technology.com/forum/index.php?topic=47427.0 Currently pre clearing drives using gfjardim's new preclear plugin / script: http://lime-technology.com/forum/index.php?topic=39985.msg453938#msg453938 BUT from the syslog it doesn't appear to the cause IMHO because what is in the log seems to start as soon as the server starts up. No disks assigned at the current time and as such Array not started. I think that you will find that is quite normal and will stop when the array is started. Link to comment
danioj Posted March 12, 2016 Share Posted March 12, 2016 On my test server, from the minute I turn it on I get a recycling set of messages for each device in the syslog over and over again like this: Mar 12 18:21:02 test kernel: sdb: sdb1 Mar 12 18:21:02 test emhttp: shcmd (1141): rmmod md-mod |& logger Mar 12 18:21:02 test kernel: md: unRAID driver removed Mar 12 18:21:02 test emhttp: shcmd (1142): modprobe md-mod super=/boot/config/super.dat |& logger Mar 12 18:21:02 test kernel: md: unRAID driver 2.6.0 installed Mar 12 18:21:02 test emhttp: Pro key detected, GUID: 0718-048A-07B2-1607419A1FD1 FILE: /boot/config/Pro.key Mar 12 18:21:02 test emhttp: Device inventory: Mar 12 18:21:02 test emhttp: shcmd (1143): udevadm settle Mar 12 18:21:02 test emhttp: TDKMedia_Gold_Flash_07B21607419A1FD1-0:0 (sda) 7811040 Mar 12 18:21:02 test emhttp: WDC_WD800AAJS-00PSA0_WD-WMAPA1162808 (sdb) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800JD-60LSA5_WD-WMAM9MF49542 (sdc) 78150712 Mar 12 18:21:02 test emhttp: ST380815AS_5RW18GXH (sdd) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800JD-60LSA5_WD-WMAM9PV29584 (sde) 78150712 Mar 12 18:21:02 test emhttp: WDC_WD800AAJS-00PSA0_WD-WMAPA1147612 (sdf) 78150712 Mar 12 18:21:02 test kernel: mdcmd (1): import 0 Mar 12 18:21:02 test kernel: md: import_slot: 0 empty Mar 12 18:21:02 test kernel: mdcmd (2): import 1 Mar 12 18:21:02 test kernel: md: import_slot: 1 missing Mar 12 18:21:02 test kernel: mdcmd (3): import 2 Mar 12 18:21:02 test kernel: mdcmd (4): import 3 Mar 12 18:21:02 test kernel: mdcmd (5): import 4 Mar 12 18:21:02 test kernel: mdcmd (6): import 5 Mar 12 18:21:02 test kernel: mdcmd (7): import 6 Mar 12 18:21:02 test kernel: mdcmd (: import 7 Mar 12 18:21:02 test kernel: mdcmd (9): import 8 Mar 12 18:21:02 test kernel: mdcmd (10): import 9 Mar 12 18:21:02 test kernel: mdcmd (11): import 10 Mar 12 18:21:02 test kernel: mdcmd (12): import 11 Mar 12 18:21:02 test kernel: mdcmd (13): import 12 Mar 12 18:21:02 test emhttp: import 30 cache device: no device Mar 12 18:21:02 test emhttp: import flash device: sda Mar 12 18:21:02 test kernel: mdcmd (14): import 13 Mar 12 18:21:02 test kernel: mdcmd (15): import 14 Mar 12 18:21:02 test kernel: mdcmd (16): import 15 Mar 12 18:21:02 test kernel: mdcmd (17): import 16 Mar 12 18:21:02 test kernel: mdcmd (18): import 17 Mar 12 18:21:02 test kernel: mdcmd (19): import 18 Mar 12 18:21:02 test kernel: mdcmd (20): import 19 Mar 12 18:21:02 test kernel: mdcmd (21): import 20 Mar 12 18:21:02 test kernel: mdcmd (22): import 21 Mar 12 18:21:02 test kernel: mdcmd (23): import 22 Mar 12 18:21:02 test kernel: mdcmd (24): import 23 Mar 12 18:21:02 test kernel: mdcmd (25): import 24 Mar 12 18:21:02 test kernel: mdcmd (26): import 25 Mar 12 18:21:02 test kernel: mdcmd (27): import 26 Mar 12 18:21:02 test kernel: mdcmd (28): import 27 Mar 12 18:21:02 test kernel: mdcmd (29): import 28 Mar 12 18:21:02 test kernel: mdcmd (30): import 29 Mar 12 18:21:02 test kernel: md: import_slot: 29 empty Mar 12 18:21:02 test emhttp: shcmd (1149): /etc/rc.d/rc.nfsd start |& logger Mar 12 18:22:37 test login[6398]: ROOT LOGIN on '/dev/tty1' Mar 12 18:23:33 test in.telnetd[29155]: connect from 192.168.1.2 (192.168.1.2) Mar 12 18:23:35 test login[29156]: ROOT LOGIN on '/dev/pts/1' from 'main.danioj.lan' To indicate what I am doing, go here: https://lime-technology.com/forum/index.php?topic=47427.0 Currently pre clearing drives using gfjardim's new preclear plugin / script: http://lime-technology.com/forum/index.php?topic=39985.msg453938#msg453938 BUT from the syslog it doesn't appear to the cause IMHO because what is in the log seems to start as soon as the server starts up. No disks assigned at the current time and as such Array not started. I think that you will find that is quite normal and will stop when the array is started. It does seem that you are correct. I have just started the Array and am undergoing a Parity Sync and the messages have stopped. Link to comment
sparklyballs Posted March 12, 2016 Share Posted March 12, 2016 changed appdata default to /mnt/cache/appdata from /mnt/user/appdata (/mnt/user/appdata just plain doesn't work for some dockers, including postgres and anything that uses sqlite) mover ran at my set time 8.00 am, and now i have an appdata folder on disk1, which is something new.... Link to comment
peter_sm Posted March 12, 2016 Share Posted March 12, 2016 VM and Qemu In the new xml file I see this line now for qemu <emulator>/usr/local/sbin/qemu</emulator> On 6.1.9 <emulator>/usr/bin/qemu-system-x86_64</emulator> Is this right ? //Peter Link to comment
danioj Posted March 12, 2016 Share Posted March 12, 2016 I'd like to know if anyone (with Dual Parity) is experiencing a dip in performance over v6.1x According to Windows 10 File Copy GUI Stats - On my test server, which I admit has some "older" disks, I don't seem to be able to get more than 26MB/s (After an initial burst of about 90MB/s for about 5 seconds) write for a 4GB file. If I enable Turbo Write I get about 60MB/s (after a similar initial burst as above) retained for the same file. To LT Staff - if you are reading - is there expected to be performance penalty as a result of implementing Dual Parity? Without Turbo Write Enabled: With Turbo Write Enabled: Link to comment
jonp Posted March 12, 2016 Share Posted March 12, 2016 VM and Qemu In the new xml file I see this line now for qemu <emulator>/usr/local/sbin/qemu</emulator> On 6.1.9 <emulator>/usr/bin/qemu-system-x86_64</emulator> Is this right ? //Peter Yes, this new wrapper script is critical to some of the upgrades to VM manager. Link to comment
jonp Posted March 12, 2016 Share Posted March 12, 2016 I'd like to know if anyone (with Dual Parity) is experiencing a dip in performance over v6.1x According to Windows 10 File Copy GUI Stats - On my test server, which I admit has some "older" disks, I don't seem to be able to get more than 26MB/s (After an initial burst of about 90MB/s for about 5 seconds) write for a 4GB file. If I enable Turbo Write I get about 60MB/s (after a similar initial burst as above) retained for the same file. To LT Staff - if you are reading - is there is there expected to be performance penalty as a result of implementing Dual Parity? Without Turbo Write Enabled: With Turbo Write Enabled: Can you really claim "slower"? Show us the same tests with the same hardware on 6.1.9 with single parity otherwise we have no baseline comparison. Link to comment
jonp Posted March 12, 2016 Share Posted March 12, 2016 changed appdata default to /mnt/cache/appdata from /mnt/user/appdata (/mnt/user/appdata just plain doesn't work for some dockers, including postgres and anything that uses sqlite) mover ran at my set time 8.00 am, and now i have an appdata folder on disk1, which is something new.... Hmm, need to see your diagnostics. Link to comment
peter_sm Posted March 12, 2016 Share Posted March 12, 2016 VM and Qemu In the new xml file I see this line now for qemu <emulator>/usr/local/sbin/qemu</emulator> On 6.1.9 <emulator>/usr/bin/qemu-system-x86_64</emulator> Is this right ? //Peter Yes, this new wrapper script is critical to some of the upgrades to VM manager. Thanks, do we still need the Qemu.conf patch for larger groups Numbers ? Link to comment
jonp Posted March 12, 2016 Share Posted March 12, 2016 VM and Qemu In the new xml file I see this line now for qemu <emulator>/usr/local/sbin/qemu</emulator> On 6.1.9 <emulator>/usr/bin/qemu-system-x86_64</emulator> Is this right ? //Peter Yes, this new wrapper script is critical to some of the upgrades to VM manager. Thanks, do we still need the Qemu.conf patch for larger groups Numbers ? Nope. Link to comment
danioj Posted March 12, 2016 Share Posted March 12, 2016 I'd like to know if anyone (with Dual Parity) is experiencing a dip in performance over v6.1x According to Windows 10 File Copy GUI Stats - On my test server, which I admit has some "older" disks, I don't seem to be able to get more than 26MB/s (After an initial burst of about 90MB/s for about 5 seconds) write for a 4GB file. If I enable Turbo Write I get about 60MB/s (after a similar initial burst as above) retained for the same file. To LT Staff - if you are reading - is there is there expected to be performance penalty as a result of implementing Dual Parity? Without Turbo Write Enabled: With Turbo Write Enabled: Can you really claim "slower"? Show us the same tests with the same hardware on 6.1.9 with single parity otherwise we have no baseline comparison. I haven't said it's slower for purely that reason (no benchmark testing was done on this hardware beforehand), therefore to everyone reading this I guess it is purely anecdotal. What I asked was: To LT Staff - if you are reading - is there is there expected to be performance penalty as a result of implementing Dual Parity? Link to comment
jonp Posted March 12, 2016 Share Posted March 12, 2016 I'd like to know if anyone (with Dual Parity) is experiencing a dip in performance over v6.1x According to Windows 10 File Copy GUI Stats - On my test server, which I admit has some "older" disks, I don't seem to be able to get more than 26MB/s (After an initial burst of about 90MB/s for about 5 seconds) write for a 4GB file. If I enable Turbo Write I get about 60MB/s (after a similar initial burst as above) retained for the same file. To LT Staff - if you are reading - is there is there expected to be performance penalty as a result of implementing Dual Parity? Without Turbo Write Enabled: With Turbo Write Enabled: Can you really claim "slower"? Show us the same tests with the same hardware on 6.1.9 with single parity otherwise we have no baseline comparison. I haven't said it's slower for purely that reason (no benchmark testing was done on this hardware beforehand), therefore to everyone reading this I guess it is purely anecdotal. What I asked was: To LT Staff - if you are reading - is there is there expected to be performance penalty as a result of implementing Dual Parity? Only in cases of extremely weak hardware (CPU). Link to comment
danioj Posted March 12, 2016 Share Posted March 12, 2016 Only in cases of extremely weak hardware (CPU). Thank You! Link to comment
binhex Posted March 12, 2016 Share Posted March 12, 2016 Hi jonp, firstly congrats on getting this out, awesome job! Do you know if the tweak to put iptable_mangle support back in got included in this release? Link to comment
jonp Posted March 12, 2016 Share Posted March 12, 2016 Hi jonp, firstly congrats on getting this out, awesome job! Do you know if the tweak to put iptable_mangle support back in got included in this release? Hi bin, Sorry for not replying to the PMs on the subject. I had forwarded internally for review. I believe this was taken care of with 6.2, but if you could confirm for the, that'd be great. Link to comment
Kimovitzh Posted March 12, 2016 Share Posted March 12, 2016 Seems nice! Any update on the Plus license storage device limits on 6.2? Link to comment
JorgeB Posted March 12, 2016 Share Posted March 12, 2016 I'd like to know if anyone (with Dual Parity) is experiencing a dip in performance over v6.1x For now only tested with SSDs, but writing to the array with single or dual parity got exactly same speed. Will test with HDDs later today. Link to comment
mauler57 Posted March 12, 2016 Share Posted March 12, 2016 When i try to boot to the gui it tries and then gives me the boot selection again. It doesnt seem to spit out an error. Is there a way to open the gui after booting the normal unraid option Link to comment
Recommended Posts