Aigor Posted November 12, 2014 Share #26 Posted November 12, 2014 Hi Hetfield, As mentioned before the problem is in a different part of the system, it has to do with all kind of checks dsm does ... checksumming ... device checking ... etc, as mentioned by schmill don't touch this until .... just my 2 cents cheers Louis as fr you know, how works checksumming? which kind of data are involved? where OS grab data? Thanks It's possibile that they put some bios module where read data for checksumming? Quote Link to comment Share on other sites More sharing options...
lowietje Posted November 12, 2014 Share #27 Posted November 12, 2014 Hi Aigor, I guess i shouldn't have replied at all But sorry to dissappoint you, i don't have any knowledge of these things, i just read the forums lately, and found some old stuff for version 4.3, i guess the checksum thing was introduced in this version .... don't pin me on this one ... So i can't help you or anyone else in this matter ... I guess this should all be done at boottime, so the nanoboot or any other boot module should be able to take care of it. Until that is done i guess there will not be a working Xpenology 5.1.... I guess we have to be patient Cheers Louis Quote Link to comment Share on other sites More sharing options...
Aigor Posted November 12, 2014 Share #28 Posted November 12, 2014 Hi lowietje and thank you for your explanation mine was "free thinking" i'm not impatient, my n40l works fine even with 5.0 DSM Quote Link to comment Share on other sites More sharing options...
yud Posted November 13, 2014 Share #29 Posted November 13, 2014 Hi yud - put simply, no. Don't touch 5.1 yet, there are various issues with it that the great minds here are still figuring out You can always keep an eye on the status table at the bottom of the page (above the comments) on xpenology.nl to see the overall status of a release: http://www.xpenology.nl/synology-released-dsm-5-1-5004/ Welcome to xpenology Thanks! I will stick to my current version for now... Quote Link to comment Share on other sites More sharing options...
Schmill Posted November 13, 2014 Share #30 Posted November 13, 2014 I don't think there is any harm in you updating to 5.0 4528 update 1 (or 2 as of today), but don't quote me on that - just don't go to 5.1 yet. Quote Link to comment Share on other sites More sharing options...
mbeckerkassel Posted November 14, 2014 Share #31 Posted November 14, 2014 This is the point i was afraid for the last weeks. we are trying to find the error, without having the option to test a solution cause 1 person got the code. i had asked here in this forum 4 weeks ago if anybody has the source for nanoboot....... no answer at all. don´t get me wrong. But open source development works differently. does anybody have the source from nanoboot? Quote Link to comment Share on other sites More sharing options...
wenlez Posted November 14, 2014 Share #32 Posted November 14, 2014 This is the point i was afraid for the last weeks. we are trying to find the error, without having the option to test a solution cause 1 person got the code. i had asked here in this forum 4 weeks ago if anybody has the source for nanoboot....... no answer at all. don´t get me wrong. But open source development works differently. does anybody have the source from nanoboot? Isn't this the source code: http://nanoboot.eu.org/source Quote Link to comment Share on other sites More sharing options...
h4rm Posted November 14, 2014 Share #33 Posted November 14, 2014 I highly recommend to put the Source of Nanoboot and Gnoboot on Github. Quote Link to comment Share on other sites More sharing options...
lehel Posted November 14, 2014 Share #34 Posted November 14, 2014 If any testing is needed let me know, also I'm happy to get into this - and debug just need some time and need to catch up whats up with XPenology, as this is my first day here, and installed 5428 currently on my HP N54L. Quote Link to comment Share on other sites More sharing options...
liwei Posted November 17, 2014 Share #35 Posted November 17, 2014 I highly recommend to put the Source of Nanoboot and Gnoboot on Github. Hey guys, please check my repo on github, it's the same kernel as nanoboot use. https://github.com/liwei/xpenology-3.x Quote Link to comment Share on other sites More sharing options...
Hennie-M Posted November 18, 2014 Share #36 Posted November 18, 2014 I highly recommend to put the Source of Nanoboot and Gnoboot on Github. Hey guys, please check my repo on github, it's the same kernel as nanoboot use. https://github.com/liwei/xpenology-3.x It is not very active, is it? Quote Link to comment Share on other sites More sharing options...
HommiePeter Posted November 19, 2014 Share #37 Posted November 19, 2014 Hey Guys, i have installed DSM 5.1 on my test system and i did boot but without my drive getting mounted als already mentioned. (See attached image) But when i did some investigation of de log files (see relevant parts of DMESG below ) i found out that: - My disk is detected - (3) Partitions where created in a Raid 1 configuration - 2 of the 3 Partitions are mounted (SWAP en System) - the 3 one (data) is not mounted because is is missing a super block Could this be the reason why the disks are disappearing, and how can we over come the missing super block issue ? Login Screen Diskstation Post init Init: Unable to mount /dev filesystem: No such device =================== start udevd ================ ===trigger device plug event ===== DiskStation login: Fragment DMESG |GREP MD [ 2.521864] md: Autodetecting RAID arrays. [ 2.557755] md: invalid raid superblock magic on sda3 [ 2.557759] md: sda3 does not have a valid v0.90 superblock, not importing! [ 2.557768] md: Scanned 3 and added 2 devices. [ 2.557769] md: autorun ... [ 2.557771] md: considering sda1 ... [ 2.557776] md: adding sda1 ... [ 2.557779] md: sda2 has different UUID to sda1 [ 2.557783] md: created md0 [ 2.557786] md: bind [ 2.557800] md: running: [ 2.557881] md/raid1:md0: active with 1 out of 12 mirrors [ 2.557894] md0: detected capacity change from 0 to 2549940224 [ 2.557929] md: considering sda2 ... [ 2.557932] md: adding sda2 ... [ 2.558072] md: created md1 [ 2.558073] md: bind [ 2.558081] md: running: [ 2.558149] md/raid1:md1: active with 1 out of 12 mirrors [ 2.558163] md1: detected capacity change from 0 to 2147418112 [ 2.558200] md: ... autorun DONE. [ 2.588148] md1: unknown partition table [ 2.619514] md0: unknown partition table [ 2.627526] EXT4-fs (md0): barriers disabled [ 2.664578] EXT4-fs (md0): mounted filesystem with ordered data mode. Opts: [ 2.755146] Adding 2097084k swap on /dev/md1. Priority:-1 extents:1 across:2097084k [ 12.564551] EXT4-fs (md0): mounted filesystem with ordered data mode. Opts: barrier=1 [ 13.059297] EXT4-fs (md0): re-mounted. Opts: (null) [ 16.126338] Adding 2097084k swap on /dev/md1. Priority:-1 extents:1 across:2097084k [ 18.187065] systemd-udevd[12176]: starting version 204 [ 19.844557] md: md2 stopped. [ 20.158934] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.336575] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.514303] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.691952] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.869715] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 21.047414] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 21.048569] md: disabled device sda3, could not read superblock. [ 21.048573] md: sda3 does not have a valid v1.2 superblock, not importing! [ 21.048580] md: md_import_device returned -22 [ 21.049702] md: md2 stopped. [ 21.213691] md: md2 stopped. Quote Link to comment Share on other sites More sharing options...
Pismo Posted November 20, 2014 Share #38 Posted November 20, 2014 Hey Guys, i have installed DSM 5.1 on my test system and i did boot but without my drive getting mounted als already mentioned. (See attached image) But when i did some investigation of de log files (see relevant parts of DMESG below ) i found out that: - My disk is detected - (3) Partitions where created in a Raid 1 configuration - 2 of the 3 Partitions are mounted (SWAP en System) - the 3 one (data) is not mounted because is is missing a super block Could this be the reason why the disks are disappearing, and how can we over come the missing super block issue ? Login Screen Diskstation Post init Init: Unable to mount /dev filesystem: No such device =================== start udevd ================ ===trigger device plug event ===== DiskStation login: Fragment DMESG |GREP MD [ 2.521864] md: Autodetecting RAID arrays. [ 2.557755] md: invalid raid superblock magic on sda3 [ 2.557759] md: sda3 does not have a valid v0.90 superblock, not importing! [ 2.557768] md: Scanned 3 and added 2 devices. [ 2.557769] md: autorun ... [ 2.557771] md: considering sda1 ... [ 2.557776] md: adding sda1 ... [ 2.557779] md: sda2 has different UUID to sda1 [ 2.557783] md: created md0 [ 2.557786] md: bind [ 2.557800] md: running: [ 2.557881] md/raid1:md0: active with 1 out of 12 mirrors [ 2.557894] md0: detected capacity change from 0 to 2549940224 [ 2.557929] md: considering sda2 ... [ 2.557932] md: adding sda2 ... [ 2.558072] md: created md1 [ 2.558073] md: bind [ 2.558081] md: running: [ 2.558149] md/raid1:md1: active with 1 out of 12 mirrors [ 2.558163] md1: detected capacity change from 0 to 2147418112 [ 2.558200] md: ... autorun DONE. [ 2.588148] md1: unknown partition table [ 2.619514] md0: unknown partition table [ 2.627526] EXT4-fs (md0): barriers disabled [ 2.664578] EXT4-fs (md0): mounted filesystem with ordered data mode. Opts: [ 2.755146] Adding 2097084k swap on /dev/md1. Priority:-1 extents:1 across:2097084k [ 12.564551] EXT4-fs (md0): mounted filesystem with ordered data mode. Opts: barrier=1 [ 13.059297] EXT4-fs (md0): re-mounted. Opts: (null) [ 16.126338] Adding 2097084k swap on /dev/md1. Priority:-1 extents:1 across:2097084k [ 18.187065] systemd-udevd[12176]: starting version 204 [ 19.844557] md: md2 stopped. [ 20.158934] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.336575] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.514303] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.691952] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 20.869715] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 21.047414] ata1.00: cmd 60/08:00:08:00:90/00:00:00:00:00/40 tag 0 ncq 4096 in [ 21.048569] md: disabled device sda3, could not read superblock. [ 21.048573] md: sda3 does not have a valid v1.2 superblock, not importing! [ 21.048580] md: md_import_device returned -22 [ 21.049702] md: md2 stopped. [ 21.213691] md: md2 stopped. Very interesting as when you create an RAID1 under Linux with mdadm command you get the following hint for the bootloader: mdadm: Note: this array has metadata at the start and may not be suitable as a boot device. If you plan to store '/boot' on this device please ensure that your boot-loader understands md/v1.x metadata, or use --metadata=0.90 So maybe nanoboot does create the RAID in a different way which isn't supported in DSM 5.1 anymore or uses an outdated version of madam. Or the kernel is too old. [root@localhost ~]# mdadm --create /dev/md0 --level=1 --raid-devices=2 /dev/sda1 /dev/sdb1 mdadm: Note: this array has metadata at the start and may not be suitable as a boot device. If you plan to store '/boot' on this device please ensure that your boot-loader understands md/v1.x metadata, or use --metadata=0.90 Continue creating array? y mdadm: Defaulting to version 1.2 metadata mdadm: array /dev/md0 started. [root@localhost ~]# More Infos can be found here for example: http://www.thomas-krenn.com/de/wiki/Linux_Software_RAID Quote Link to comment Share on other sites More sharing options...
Poechi Posted November 20, 2014 Share #39 Posted November 20, 2014 (edited) - Edited July 10, 2015 by Guest Quote Link to comment Share on other sites More sharing options...
Giorgio Posted November 20, 2014 Share #40 Posted November 20, 2014 So, I think it safe to say DSM 5.1 uses superblock 1.2 for the data array. Correct, but as DSM 5.0-4528 U2 also uses superblock 1.2 it can't be the cause of problem which we are having with DSM 5.1. I installed 5.1 on a simple bare metal test system and I see the following: Starting /usr/syno/bin/syncfgen... [OK] /usr/syno/bin/synocfgen returns 0 Partition Version=0 Partition layout is not DiskStation style. NOT EXECUTE /sbin/e2fsck. Mounting /dev/md0 /tmpRoot :: Checking upgrade file [OK] linuxrc.syno executed successfully. Post init init: Unable to mount /dev filesystem: No such device =================== start udevd ================ ===trigger device plug event ===== DiskStation login: Quote Link to comment Share on other sites More sharing options...
Poechi Posted November 20, 2014 Share #41 Posted November 20, 2014 (edited) - Edited July 10, 2015 by Guest Quote Link to comment Share on other sites More sharing options...
Pismo Posted November 20, 2014 Share #42 Posted November 20, 2014 Whatever "Partition layout is not DiskStation style." means. Quote Link to comment Share on other sites More sharing options...
jagwaugh Posted November 21, 2014 Share #43 Posted November 21, 2014 It isn't the superblock on the partition, it has been 1.2 for a while. The seems that the volume problem is related to device detection. If you upgrade a DSM5.0 to DSM5.1 md0 and md1 are preserved and work correctly, it is only md2 (where the volume is) that stops working. The volume and data are all there though - if you boot to a different linux, and assemble the raids you will find that all 3 work fine. I think the problem is device detection during the boot: The boot message "init: Unable to mount /dev filesystem: No such device" is one clue, the other is that if you try "mdadm --assemble --scan" it tries to assemble md2 out of "8;53, 8:69, 8:85, and 8:37 when it should be assembling the array out of the sd[abcd]5 partitions. "cat /proc/partitions" indicates that sd[abcd]5 exist, but "ls /dev/sd*" indicates there are no sdX devices, while "ls /dev/hd* shows that the drives are apparently all there, but detected as hdX devices. Trying to assemble md2 from the hdX5 partitions doesn't work either. Googling around for the "init: Unable to mount /dev filesystem: No such device" and the misdetection of sdX as hdX it would seem that CONFIG_DEVTMPFS was not set when the xpenology boot image was compiled, but this may be a red herring. liwei's Git repository has this flag set. Nonetheless, I think it is related to the device detection. I am trying to set up a VM to compile the kernel but I'm not having much luck - on the one hand, I'm not all that proficient at linux, and on the other, I think we need the gpl for DSM5.1 from synology, which hasn't been released yet. Has anyone seen a complete "how to set up for cross compiling xpenology guide"? The "partition layout is not diskstation style" has been there since quite a few versions of xpenology, so I don't think that is the issue. Quote Link to comment Share on other sites More sharing options...
LeaderGL Posted November 21, 2014 Share #44 Posted November 21, 2014 Has anyone seen a complete "how to set up for cross compiling xpenology guide"? try with: 1. http://hallard.me/how-to-install-kernel-modules-on-synology-ds1010-dsm-4-1/ 2. http://xpenology.com/wiki/en/building_xpenology Quote Link to comment Share on other sites More sharing options...
Poechi Posted November 21, 2014 Share #45 Posted November 21, 2014 (edited) - Edited July 10, 2015 by Guest Quote Link to comment Share on other sites More sharing options...
HommiePeter Posted November 21, 2014 Share #46 Posted November 21, 2014 Hi there all again, i have been looking more into the issue of the volumes not getting mounted. I have seen 2 points of interest the seem to be linked to each other. 1) DSM 5.1 uses AppArmor 2) the boot script has been seriously changed for mounting the volumes DSM 5.1 uses AppArmor. DSM uses AppArmor for high security protection. I expects that the relevant modules are load in the linux kernel. De kernel that is used for Nanoboot does not support AppArmor which leads to faults during booting. I am not sure if activation AppArmor in the Nanoboot linuxkernel requires a complete rebuild of the kernel ...... As an alternative i am looking at Disabling AppArmor in /etc/synoinfo.conf and /etc.defaults/synoinfo.conf ... support_apparmor="yes" Running /usr/syno/etc/rc.sysv/apparmor.sh status Gives AppArmor is not loaded. Install the apparmor-utils package to receive more detailed status information here (or examine directly). Different boot script. In DSM 5.1 the boot script has been changed (etc/rc) especially on mounting of the volumes. See included code examples..... Part of the old bootscript DSM 5.0 4493 on a DS214play. if [ "yes" = "$RUN_HA" ]; then $SZF_HA_RC prepare-for-upg else # initial findhostd first to report quota check progress, see DS20 bug # /usr/syno/etc/rc.d/S98findhostd.sh start if [ "yes" = "${UbifsVolume}" ]; then /bin/mount -t ubifs ubi0:volume1 /volume1 else # checking and Mounting filesystem(s) ... /usr/syno/bin/synobootseq --set-check-fs >/dev/null 2>&1 /etc.defaults/rc.volume start # reset bootseq to start-services /usr/syno/bin/synobootseq --set-start-services >/dev/null 2>&1 fi /sbin/initctl emit --no-wait syno.volume.ready fi Part of the new bootscript on DSM 5.1 5004 on a DS3612xs. if [ "yes" = "$RUN_HA" ]; then $SZF_HA_RC prepare-for-upg else # checking and Mounting filesystem(s) ... /usr/syno/bin/synobootseq --set-check-fs >/dev/null 2>&1 /etc.defaults/rc.volume start # reset bootseq to start-services /usr/syno/bin/synobootseq --set-start-services >/dev/null 2>&1 /sbin/initctl emit --no-wait syno.volume.ready fi When manually running the steps on DSM 5.1 /usr/syno/bin/synobootseq --set-check-fs >/dev/null 2>&1 /etc/rc.volume start /usr/syno/bin/synobootseq --set-start-services >/dev/null 2>&1 /sbin/initctl emit --no-wait syno.volume.ready I get the following output which links it back to the AppArmor module which is not loaded (see point 1) DiskStation> Starting AppArmor modprobe: chdir(3.2.40): No such file or directory $Loading AppArmor module: Failed. I know this gives not the solution we all are looking for ...... but might provide some more insight the the problem we need to tackle.... Quote Link to comment Share on other sites More sharing options...
jagwaugh Posted November 22, 2014 Share #47 Posted November 22, 2014 So, I have setup a Ubuntu server vm, latest updates, and tried using sancome's git repository (and scripts) and the other files per nanoboot.eu I ran: 01-unzip-linux-3.x.shmake clean make menuconfig without seeing any error messages. When I run: make modules It outputs: fs/nfsd/vfs.c: In function ‘nfsd_setattr’: fs/nfsd/vfs.c:337:2: error: implicit declaration of function ‘IS_SYNOACL’ [-Werror=implicit-function-declaration] cc1: some warnings being treated as errors make[2]: *** [fs/nfsd/vfs.o] Error 1 make[1]: *** [fs/nfsd] Error 2 make: *** [fs] Error 2 This makes make bzImage barf, even if I don't change anything in menuconfig. Question 1) Why is make modules giving this error? Question 2) Even if I do get this to compile an image, this will be for DSM50, don't I need gpl and toolchains for DSM5.1? Andrew ps. As far as I can tell, CONFIG_DEVTMPFS is set in Sancome's .config, I was hoping that it wasn't, as this can cause device detection issues similar to what we have been seeing on DSM5.1. Quote Link to comment Share on other sites More sharing options...
Guest Posted November 24, 2014 Share #48 Posted November 24, 2014 is it possible to disable completly apparmor ? perhaps, it's solve the problem ?? /usr/syno/etc.defaults/rc.sysv/apparmor.sh stop Quote Link to comment Share on other sites More sharing options...
djelusion Posted November 26, 2014 Share #49 Posted November 26, 2014 so in the 5 pages of comments i take it nobody was successful in getting this to work on any hardware. That is kind of scary to think about it, since nobody has this working and its been a good couple of weeks or so. I wonder if by this year in 2014 will it be possible to run 5.1 or have the xpenology box not be accessible from the internet. Quote Link to comment Share on other sites More sharing options...
Vortex Posted November 26, 2014 Share #50 Posted November 26, 2014 Strange that Synology still have not posted the new kernel sources. They violates the GPL policy again! Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.