Degraded Raid1 with healthy drives on DNS-323 Rev C with 0.1RC5, not accessible anymore with Samba

[Tibo]

Hello,

I've been having a degraded Raid1 configuration for a while now (months/years). I have 2 identical 2TB drives and they look healthy. I didn't have any issue with the NAS before, except for very slow reboot/resync on hard reset (power outtage).

I had another power outtage recently (the joys of canadian winters) and since then, the device boots up, but is not accessible via samba (shared drive). Happened on 0.1RC4, upgraded to RC5, no success.

Power button and both drive LEDs are amber, Lan LED is blue.

I realized I had issues with the fan1_input, but reading the forum, I fixed the simlink. Though in my case, symlink was trying to access

/sys/class/hwmon/hwmon1/fan1_input

(not hwmon0), but I pointed to 

/sys/class/hwmon/hwmon0/device/fan1_input

Not sure if the hwmon0 vs. hwmon1 would cause an issue or is completely unrelated.

Anyways, after running rcsysctrl, I end up with the following error when trying to access via samba:

Feb  4 00:15:54 dlink-NAS daemon.info sysctrl: signaled to quit, quiting
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: Starting
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: reading /etc/bay
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: reading /etc/misc.conf
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: reading /etc/sysctrl.conf
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: No /etc/sysctrl.conf: No such file or directory Using defaults
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.lo_fan=2000
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.hi_fan=5000
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.lo_temp=45
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.hi_temp=50
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.hist_temp=2.0
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.fan_mode=AUTO
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.mail=1
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.recovery=1
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.fan_off_temp=38
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.max_fan_speed=6000
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.crit_temp=54
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.warn_temp=52
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.crit_temp_command="/usr/sbin/poweroff"
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.warn_temp_command="(null)"
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.front_button_command1="(null)"
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.front_button_command2="(null)"
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: args.back_button_command="(null)"
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: ndisks=2
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: sdb left rdwr=0 last=0 spindow=1200 power=1
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: sda right rdwr=0 last=0 spindow=1200 power=1
Feb  4 00:15:55 dlink-NAS daemon.info sysctrl: (null) (null) rdwr=0 last=0 spindow=1200 power=1
Feb  4 00:15:55 dlink-NAS daemon.crit sysctrl: Your md0 raid1 device is degraded!

BTW, I can navigate the drive content from the web interface (Setup > Folders) or from SSH without issue.

Any help really appreciated!

[João Cardoso]

On Saturday, 4 February 2017 05:33:29 UTC, Tibo wrote:

Hello,

I've been having a degraded Raid1 configuration for a while now (months/years). I have 2 identical 2TB drives and they look healthy.

You can try re-adding the missing component to the RAID, see the recent degraded RAID5 topic (procedure is identical for a RAID1)

As the missing component has been offline for months it will certainly require a long resync.

I didn't have any issue with the NAS before, except for very slow reboot/resync on hard reset (power outtage).

You should convert your ext2 filesystems to ext3 or ext4 (Disk->Filesystem), otherwise a lengthy check will always be done after a power cut.

Converting is a very long lasting operation.

 

I had another power outtage recently (the joys of canadian winters) and since then, the device boots up, but is not accessible via samba (shared drive). Happened on 0.1RC4, upgraded to RC5, no success.

Does the shares and user accounts still exists? Can users still login with ftp or ssh? It can be configuration files corruption.

See Setup->Users and verify that/if normal users can login using ftp or ssh (then the user account exists).

 

see also Services->Network, smb, Configure. From the command line you can try 'smbtree -N' to try browsing the network; the box itself should appear with its defined shares.

Power button and both drive LEDs are amber,

Meaning that an error/warning appears in the Status page

 

Lan LED is blue.

I realized I had issues with the fan1_input, but reading the forum, I fixed the simlink. Though in my case, symlink was trying to access

/sys/class/hwmon/hwmon1/fan1_input

(not hwmon0), but I pointed to 

/sys/class/hwmon/hwmon0/device/fan1_input

Not sure if the hwmon0 vs. hwmon1 would cause an issue or is completely unrelated.

Different boxes model/revisions have the real files in different locations, so it can be hwmon0 or hwmon1

 

Anyways, after running rcsysctrl, I end up with the following error

 

Those are not error, they are info, look at the  "daemon.info"; the last line yes, its an error and critical: "daemon.crit"

[Tibo]

Hello,

Thanks a lot for the answer!

I've been trying to understand and find the time to try your answer. I'm still trying to understand the thread which I think you're refeing to (this one?). I'm not even sure if the drive (sda) is dettached. It does only list sdb2 in Disk > RAID under RAID Maintenance (I had attached the screenshot in the first post).

While I continue, here are some more details:

1- NFS works just well, tried it on my linux laptop, really fast, no issue.

2- SSH works well too

3- FTP works, but is pretty slow (compared to samba and NFS)

4- Permissions seem good (the group is set with R/W permissions, my users are all part of that group)

In Services > Network, I can browse the files.

running >smbtree -N using SSH, I got:

$ smbtree -N

Receiving SMB: Server stopped responding

anonymous failed session setup with NT_STATUS_IO_TIMEOUT

Then it listed my network with my shared drives, including the shares on the NAS.

Strangely, I seemed to have an error with one of my users, but it's not the user I usually use for shares. I deleted the user and re-created it, the new one works fine. (actually the old one has an uppercase first letter, not sure it's considered the same as the all lowercase one)

Now, I was wondering: can I change the filesystem without impacting the content? Or should I make sure to backup everything? My issue is I have around 800Go worth of backup (years of photos and backups)...

Thanks!

[João Cardoso]

On Friday, 10 February 2017 04:34:29 UTC, Tibo wrote:

Hello,

Thanks a lot for the answer!

I've been trying to understand and find the time to try your answer. I'm still trying to understand the thread which I think you're refeing to (this one?).

No, it was https://groups.google.com/d/msg/alt-f/DOOcl5ZZnqs/n0moat0lBgAJ

You only have to select 'sda2' and 'Add' it to the RAID using the RAID webUI, Component Operations.

 

I'm not even sure if the drive (sda) is dettached. It does only list sdb2 in Disk > RAID under RAID Maintenance (I had attached the screenshot in the first post).

 

While I continue, here are some more details:

1- NFS works just well, tried it on my linux laptop, really fast, no issue.

2- SSH works well too

3- FTP works, but is pretty slow (compared to samba and NFS)

Something is not OK, ftp is as fast or faster then samba and nfs. You client must be requesting sftp or ftps, which uses encryption and is much slower  than plain ftp.

 

4- Permissions seem good (the group is set with R/W permissions, my users are all part of that group)

In Services > Network, I can browse the files.

running >smbtree -N using SSH, I got:

$ smbtree -N

Receiving SMB: Server stopped responding

anonymous failed session setup with NT_STATUS_IO_TIMEOUT

Then it listed my network with my shared drives, including the shares on the NAS.

 

Better use a separated thread for that.

Strangely, I seemed to have an error with one of my users, but it's not the user I usually use for shares. I deleted the user and re-created it, the new one works fine. (actually the old one has an uppercase first letter, not sure it's considered the same as the all lowercase one)

Now, I was wondering: can I change the filesystem without impacting the content? Or should I make sure to backup everything?

The conversion from ext2 to ext3 or ext4 preserves the data, but having a backup is always the recommended procedure

 

My issue is I have around 800Go worth of backup (years of photos and backups)...

I understand that, I still have ext3 on my own (years old) Media files.

But at the time I converted from ext2 to ext3 I backup it first using a 1TB WD Passport attached drive -- its's always handy to have one

[Thibaud Lescuyer]

Thanks again!

I started to copy all my files to a drive on my main computer, should have enough space.

I'm using FTP, transfer rate seems pretty fast (2 files at the same time at around  10 MB/s, so pretty much what I had with samba). I think the slowness was when I tried to export my pictures directly to the NAS using FTP, so it might have been an issue with the software's FTP client, or maybe I was running some other tests on the NAS, or I set the wrong port ans used FTPS, can't remember.

I started a long SMART test on the sda drive, should be done by tomorrow, the backup should be done too at that time, if everything's alright, I'll try to re-attach the drive, then convert to ext4.

I'll keep you posted!

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to alt-f+un...@googlegroups.com.
Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

[Thibaud Lescuyer]

Hello,

When I clicked Add, I got the following error:

I'm thinking about clearing the sda drive, then adding it, not sure if that's the right way.

Everything's backed-up on my computer.

What do you think?

[João Cardoso]

On Sunday, 12 February 2017 16:44:29 UTC, Tibo wrote:

Hello,

When I clicked Add, I got the following error:

I'm thinking about clearing the sda drive, then adding it, not sure if that's the right way.

Everything's backed-up on my computer.

in the topic I refer to above:

You can try to "Add" it again to the RAID (Disk->RAID "Component Operations"), of use "Clear" then "Add" if Add failed the first time.

and latter

I see no other indication of troubles, so I think that's OK to Add it to the RAID (and wait for a lengthy tens of hours rebuild). As said, if it fails the first Add, you have to Clear it first before trying again.

In that topic, adding succeed the first time and resync was fast, because the partition was not offline from the RAID for a long time. In your case, months in the degraded state, (re) adding failed and you have to clear the RAID info from the partition and a full copy will be performed.

 

What do you think? 

On Fri, Feb 10, 2017 at 5:05 PM Thibaud Lescuyer wrote:

Thanks again!

I started to copy all my files to a drive on my main computer, should have enough space.

I'm using FTP, transfer rate seems pretty fast (2 files at the same time at around  10 MB/s, so pretty much what I had with samba). I think the slowness was when I tried to export my pictures directly to the NAS using FTP, so it might have been an issue with the software's FTP client, or maybe I was running some other tests on the NAS, or I set the wrong port ans used FTPS, can't remember.

I started a long SMART test on the sda drive, should be done by tomorrow, the backup should be done too at that time, if everything's alright, I'll try to re-attach the drive, then convert to ext4.

I'll keep you posted!

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

[Tibo]

Sorry, should have re-read the thread when I had the issue.

So I did a Clear, but then after the sda was not available to add. The disk wasn't mounted (saw that in Disk > Filesystems), I had to do a "check" on it, which I did, everything looks good and the disk was automatically mounted after the check.

Issue is the sda disk is still not available in the components of the RAID, so I can't mount it (see screenshots).

Any idea?

To unsubscribe from this group and all its topics, send an email to alt-f+un...@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

[João Cardoso]

On Monday, 13 February 2017 03:46:58 UTC, Tibo wrote:

Sorry, should have re-read the thread when I had the issue.

So I did a Clear, but then after the sda was not available to add. The disk wasn't mounted (saw that in Disk > Filesystems), I had to do a "check" on it, which I did, everything looks good and the disk was automatically mounted after the check.

Issue is the sda disk is still not available in the components of the RAID, so I can't mount it (see screenshots).

Any idea?

Yes. You have a D-Link inherited RAID setup.

Nothing wrong with that, but D-Link don't mark each disk partition with its intended finality, and on the RAID webUI Alt-F only shows disk partitions intended to be used for RAID (or disk partitions that are currently used as a RAID partition).

So you have to go to Disk->Partitioner, select the sda disk on the upper section, and then uncheck the "Keep" checkbutton on the "sda2" partition, and change the "Type" to RAID **WITHOUT** changing anything else, then hit the "Partition" button.

After that "sda2" will appear as a possible component on the RAID page.

Notice also that "sda2" is a partition, a disk has several partitions on it, the RAID is made up using disk partitions, not the whole disk, and a partition can be used for other purposes, such as holding a filesystem which manages your data. All that is tersely explained in the Disk Partitioner, RAID setup, Disk Filesystem online help (the blue (?) icon near the pages title).

So, you didn't need to check and mount sda2, which is now a plain filesystem that holds the data that was on the RAID1 when it become degraded.

[Thibaud Lescuyer]

Awesome! Will try that as soon as possible (maybe tomorrow night, Eastern Time). Will keep you posted!

Again, thanks a lot for the help, that's really appreciated!

[Tibo]

Hi,

I got the following error (disk-partitioner-sda.png) and clicking OK does not seem to do anything.

Reading the Disk Partitioner page, I understand that the start sector and length of some partitions might cause issues. As they were created with D-Link, I didn't get that error/fix from ALT-F (about 8-multiple blocks).

From what I understand, my options would be to:

1- manually change the start and lengh of the faulty partitions (by shrinking only to prevent data loss, loosing some small disk space between partitions)

or

2- re-do the whole RAID setup

As you can see in the second screenshot (about sdb disk), the partition table is exactly the same.

What do you recommend?

Thanks again ☺

On Monday, February 13, 2017 at 2:43:50 PM UTC-5, Tibo wrote:

Awesome! Will try that as soon as possible (maybe tomorrow night, Eastern Time). Will keep you posted!

Again, thanks a lot for the help, that's really appreciated!

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

[João Cardoso]

On Wednesday, 15 February 2017 04:07:03 UTC, Tibo wrote:

Hi,

I got the following error (disk-partitioner-sda.png) and clicking OK does not seem to do anything.

Yes, D-Link didn't know at the time about 4KB sector disks, so its partition scheme didn't take that into account. Having partitions not 4K aligned means a lower writing speed.

Reading the Disk Partitioner page, I understand that the start sector and length of some partitions might cause issues. As they were created with D-Link, I didn't get that error/fix from ALT-F (about 8-multiple blocks).

From what I understand, my options would be to:

1- manually change the start and lengh of the faulty partitions (by shrinking only to prevent data loss, loosing some small disk space between partitions)

Changing the start sector number of a partition means loosing all its existing data. Changing only the partition length might lead to some data loss if the filesystem is not previously prepared for that.

The above has no implications on sda2, as its data is not relevant as it will be overwritten when it will be added to the RAID (sdb2 will be copied over sda2).

But it will have implications on sda4 -- are you using it for any purpose? Perhaps using it for Alt-F packages?

 

or

2- re-do the whole RAID setup

That is a possibility, but if you have Alt-F packages installed they will all be removed.

 

As you can see in the second screenshot (about sdb disk), the partition table is exactly the same.

What do you recommend?

If you are satisfied with your current setup, the faster track is to change sda2 partition type using the command line (unmount first sda2)

sfdisk -c /dev/sda 2 da # change sda second disk partition to type 'da'
sfdisk-R /dev/sda # make kernel reread partition table

Now sda2 will appear as a possible RAID component on the RAID webUI and you can 'Add' it to the RAID1 (doing that from the command line is also simple).

But its partition table will remain not 4K aligned. And as sdb2 is identical and is also not 4K aligned the whole RAID suffers from write performance. It might be negligible or be 10-30% lower than it could be.

The only cure is to use the Disk Wizard on both disks to create new partitions, RAID1 and filesystem, and then copy all backup data from the PC to the new RAID, which will take a while. And re-installs Alt-F packages.

So, it all depends on your real needs and evaluation of the RAID1 previous performance (do you write to it often?) and the time you can spare to do the changes.

[Thibaud Lescuyer]

Hello,

Thank you very much once again. I've always find the write (and/or read?) to be pretty slow, and as I regularly transfer photos to the NAS, and given that I've already backed-up everything and that SAMBA is down, I think I'll try to re-format everything using the wizard.

I haven't installed any ALT-F packages.

Are you the only/main dev of ALT-F ?

[Tibo]

So here what I ended up doing:

1- Tried to re-create the RAID1 setup using the wizard

I kept having this error:

eject error, st=0

 

Reading other threads, I did the following to fix it.

2- Fix the eject error

1. Download DNS's original firmware on their support site
2. Flashed the original firmware
3. From the DNS firmware, flashed ALT-F RC5

3- Use the wizard to create my RAID1 setup

From there, the wizard worked and I was able to re-crete the RAID1.

After around 16 hours of resyncing, the RAID is ready as of this morning.

4- Configure SAMBA/NFS and transfer backup

When I configured my users and Samba and NFS services, the shared drive was now available on the network, as it was before.

I'm now transfering back my 600GB of data backed-up on my computer. Should be in around 18h to 56 days :) (I have lots of small files, so the transfer is pretty slow but goes up to 14MB/s for bigger files. ).

Thanks a lot João! Your long help and explanation was extremely helpful and I learned a lot. Plus I discovered how great ALT-F is!

I'd like to make a contribution, is donating on the project's Google Sites page something that would help you personnaly?

Thanks again ☺

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

[João Cardoso]

On Sunday, 19 February 2017 15:19:48 UTC, Tibo wrote:

So here what I ended up doing:

1- Tried to re-create the RAID1 setup using the wizard

I kept having this error:

eject error, st=0

That is an elusive bug that I can't trace back.

I think that it could be solved by using Packages->Alt-F, unchecking the "Boot Enabled" checkbox in the "Packages Installed On" section, then reboot.

 

 

Reading other threads, I did the following to fix it.

2- Fix the eject error

1. Download DNS's original firmware on their support site
2. Flashed the original firmware

that formatted the drives, "solving" the problem.

 

1. From the DNS firmware, flashed ALT-F RC5

Don't forget to apply the top-posted "RC5 network corruption" fix

 

3- Use the wizard to create my RAID1 setup

From there, the wizard worked and I was able to re-crete the RAID1.

After around 16 hours of resyncing, the RAID is ready as of this morning.

4- Configure SAMBA/NFS and transfer backup

When I configured my users and Samba and NFS services, the shared drive was now available on the network, as it was before.

I'm now transfering back my 600GB of data backed-up on my computer. Should be in around 18h to 56 days :) (I have lots of small files, so the transfer is pretty slow but goes up to 14MB/s for bigger files. ).

Thanks a lot João! Your long help and explanation was extremely helpful and I learned a lot. Plus I discovered how great ALT-F is!

I'd like to make a contribution, is donating on the project's Google Sites page something that would help you personnaly?

It's always welcome :-)

On Friday, February 17, 2017 at 8:49:37 AM UTC-5, Tibo wrote:

Hello,

Thank you very much once again. I've always find the write (and/or read?) to be pretty slow, and as I regularly transfer photos to the NAS, and given that I've already backed-up everything and that SAMBA is down, I think I'll try to re-format everything using the wizard.

I haven't installed any ALT-F packages.

Are you the only/main dev of ALT-F ?

[Thibaud Lescuyer]

Did it really format the drives? I think once I had re-installed ALT-F, I could see my RAID drive (the degraded one). Not 100% sure.

Anyways, I couldn't fix my issue, but at least I could get back to a functioning setup.

I haven't seen any big improvement in terms of throughput, still stuck as 12MB/s max. I boosted the frame size at 3500 (setting it at 7500 rendered my NAS inaccessible, had to reset it). Activated 9000 support on my desktop and everything is setup in 1Gbps, double-checked it. But that's the best throughput I already had, so it's OK I guess. Might be a little more responsive, not sure.

Thanks again for the help :)

[João Cardoso]

On Monday, 20 February 2017 12:46:44 UTC, Tibo wrote:

Did it really format the drives? I think once I had re-installed ALT-F, I could see my RAID drive (the degraded one). Not 100% sure.

It might depends on the fw version, it used to format the drives if it didn't recognized the layout

 

Anyways, I couldn't fix my issue, but at least I could get back to a functioning setup.

I haven't seen any big improvement in terms of throughput, still stuck as 12MB/s max.

Reading speed should not be affected, but writing will benefice from the right partition alignment. But remember that now the RAID is not degraded anymore, so that are two full writes, one for each disk, so you might not notice write speed improvements (but you have gained redundancy, so the overall sum is not zero)

 

I boosted the frame size at 3500 (setting it at 7500 rendered my NAS inaccessible, had to reset it). Activated 9000 support on my desktop and everything is setup in 1Gbps, double-checked it. But that's the best throughput I already had, so it's OK I guess.

That's the average. For large files, using jumbo frames really helps (but that's not a panacea)

 

Might be a little more responsive, not sure.

Thanks again for the help :)

Thanks again ☺

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

[Thibaud Lescuyer]

Again, thanks for the info!

I... I'm completely out of words... I tried to manage my pictures from the NAS, as I usually do and I just discoverd that most files (as in close to 99%) are corrupted.

The backup I had on my computer has the same corruption, so it happened when I transfered from the degraded NAS to my computer.

I just lost all my backup, and 10 years of pictures...

A lot of the pictures are already backuped up on Google Photos, but I lost everything else..

That's around 800GB of lost data...

I can read all the folders without issues, but most files are corrupted which prevent a correct reading. Pictures display colored stripes, PDFs just stop rendering...

I'm completely dumbfounded...

As I've already re-wrote everything back on the drives, I guess I can't recover the files (as when you loose the partition table, the files are still in there).

:(

[João Cardoso]

On Tuesday, 21 February 2017 03:49:27 UTC, Tibo wrote:

Again, thanks for the info!

I... I'm completely out of words... I tried to manage my pictures from the NAS, as I usually do and I just discoverd that most files (as in close to 99%) are corrupted.

The backup I had on my computer has the same corruption, so it happened when I transfered from the degraded NAS to my computer.

I just lost all my backup, and 10 years of pictures...

A lot of the pictures are already backuped up on Google Photos, but I lost everything else..

That's around 800GB of lost data...

I can read all the folders without issues, but most files are corrupted which prevent a correct reading. Pictures display colored stripes, PDFs just stop rendering...

I'm completely dumbfounded...

As I've already re-wrote everything back on the drives, I guess I can't recover the files (as when you loose the partition table, the files are still in there).

:(

I'm really very sorry about your lost photos, and I'm afraid there is nothing one can do.

It seems that you have not applied the "RC5 network corruption" fix, and the backup you made to your PC went corrupted.

As you started this topic reporting that you had issues with the system temperature reading and fixed it by reading the forum, I wrongly assumed that you have also read the network correction topic and applied the fix. Even so, three posts above I reminded you to re-apply the fix after re formatting the drives, but at that time it was already too late.

In my defense, the issue warning is displayed in the project Download webpage, which displays the README-0.1RC5.txt file, appears in the project News, appears in the top-posted RC5 announcement and WARNING: Network data corruption on RC5 topics at the forum, and appeared in the News section of the box Status webUI until you cleared it (possibly it was intermixed with packages updates and other announcements).

I'm really very sorry, feel distressed and I'm questioning myself about Alt-F future.

But you should feel worse.

João

Thanks again ☺

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

[Thibaud Lescuyer]

Hello,

Again, nothing about you, I had read about the fix (later) and applied it on the new setup, but I'm pretty sure I did forget / didn't know that I needed to apply it before.

I should have double checked the backup myself first before re-formatting the NAS drives.

It's alright, most of my pictures were already exported as compressed JPGs and uploaded on Google Photos and my phone pictures have been backed up on Photos for years.

This might give me some frustrations from time to time in the future (when I would have liked to access/use some of the files I had there), but I really don't think it's that big of an issue.

The main point of the NAS for me was for backup. My current computer is still UP, so I haven't lost the documents that are here.

I think I only really lost the pictures I've taken since January (pictures I hadn't treated/exported to Google Photos), so it's only some personal pictures that I took for fun. They might still be on my SD card, or recoverable from there, I'll check.

Don't worry, I shared my feeling yesterday as I was still in "shock", but I think it's not that bad :)

I'll make sure to have an external backup (online) for some critical data, as, if my house burns, I'm not covered, but I was already ready to loose my pictures in that case. So it's OK :)

I realized with the multiple ALT-F packages (never heard of them before you mentioned them) that I could use the NAS for media management, sharing and player (UPNP/dlna). This is a use case that I guess many users love and I'm actually thinking about it for someone I know who has issues with his current setup. But for me, the most important part was have my pictures secured in redundancy, and I failed it with a mis manipulation.

The error is all mine and again, I thank you deeply for all the work on ALT-F and on helping me (and many others as I've seen in the different posts on the forum).

Don't worry, everything will be fine :)

On Tue, Feb 21, 2017 at 1:38 PM João Cardoso <whoami...@gmail.com> wrote:

On Tuesday, 21 February 2017 03:49:27 UTC, Tibo wrote:

Again, thanks for the info!

I... I'm completely out of words... I tried to manage my pictures from the NAS, as I usually do and I just discoverd that most files (as in close to 99%) are corrupted.

The backup I had on my computer has the same corruption, so it happened when I transfered from the degraded NAS to my computer.

I just lost all my backup, and 10 years of pictures...

A lot of the pictures are already backuped up on Google Photos, but I lost everything else..

That's around 800GB of lost data...

I can read all the folders without issues, but most files are corrupted which prevent a correct reading. Pictures display colored stripes, PDFs just stop rendering...

I'm completely dumbfounded...

As I've already re-wrote everything back on the drives, I guess I can't recover the files (as when you loose the partition table, the files are still in there).

:(

I'm really very sorry about your lost photos, and I'm afraid there is nothing one can do.

It seems that you have not applied the "RC5 network corruption" fix, and the backup you made to your PC went corrupted.

As you started this topic reporting that you had issues with the system temperature reading and fixed it by reading the forum, I wrongly assumed that you have also read the network correction topic and applied the fix. Even so, three posts above I reminded you to re-apply the fix after re formatting the drives, but at that time it was already too late.

In my defense, the issue warning is displayed in the project Download webpage, which displays the README-0.1RC5.txt file, appears in the project News, appears in the top-posted RC5 announcement and WARNING: Network data corruption on RC5 topics at the forum, and appeared in the News section of the box Status webUI until you cleared it (possibly it was intermixed with packages updates and other announcements).

I'm really very sorry, feel distressed and I'm questioning myself about Alt-F future.

[Tom Schmidt]

João, would it be possible to release a 0.1RC5.1 or 0.1RC6 that incorporates the fix for the data corruption and other workarounds that have needed to be applied to the 0.1RC5 release?  It would be better if a stable pre-patched release was the default current release to download, rather than the users having to patch it immediately or possibly suffer data corruption.

Tom

Thanks again ☺

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--
You received this message because you are subscribed to a topic in the Google Groups "Alt-F" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/alt-f/bR6apXyl_Mw/unsubscribe.

To unsubscribe from this group and all its topics, send an email to alt-f+unsubscribe@googlegroups.com.

Visit this group at https://groups.google.com/group/alt-f.
For more options, visit https://groups.google.com/d/optout.

--

Thibaud Lescuyer
Développeur Web
Cell: 514-836-7886
http://ca.linkedin.com/in/thibaudlescuyer

--

You received this message because you are subscribed to the Google Groups "Alt-F" group.
To unsubscribe from this group and stop receiving emails from it, send an email to alt-f+unsubscribe@googlegroups.com.

[João Cardoso]

On Tuesday, 21 February 2017 19:31:47 UTC, Tom Schmidt wrote:

João, would it be possible to release a 0.1RC5.1 or 0.1RC6 that incorporates the fix for the data corruption and other workarounds that have needed to be applied to the 0.1RC5 release?

I have considered that, but it was not possible. The issue is in the linux kernel itself, and I couldn't find a backport. I'm currently using the last longterm stable kernel 4.4.45 (three days ago, now it is at 4.4.50...), and it still hasn't the bug fixed, I have to apply a patch to that.

The fix thus requires the ethtool package, and that can't be included in the DNS-323/321 fw because of flash memory constrains. It has to be installed on disk as a package.

I could release a "fixup", also disk-resident, but that is difficult to do, as users can have its own written boot script, and conflicts could arise; and users would have to apply it anyway, so that is not very different than following the 5 step fix instructions.

Another possibility would be to use some heuristics to warn the user that a fix needs to be applied, but again that depends on the user, as there are no open back-doors to do that otherway.

I could just remove RC5, but some packages would have to be rebuild mainly because of openssl drop on SSL2/3. And RC5 release and bug discovery just happened before my *long* vacations, so I had no time left.

So I think that I pondered all options and decided to leave it as is to not increase entropy/confusion, documenting and alerting users for the bug  on several ways.

I'm currently svn committing changes and I'm thinking to make a RC6-snapshot available, so users can test it, particularly for the boxes which I don't have and I can't test it myself (DNS-323-rev-C, 321-rev-xx, 320-rev-A)

 

  It would be better if a stable pre-patched release was the default current release to download, rather than the users having to patch it immediately or possibly suffer data corruption.

Tom

On Tue, Feb 21, 2017 at 11:51 AM, Thibaud Lescuyer  wrote:

Hello,

Again, nothing about you, I had read about the fix (later) and applied it on the new setup, but I'm pretty sure I did forget / didn't know that I needed to apply it before.

I should have double checked the backup myself first before re-formatting the NAS drives.

It's alright, most of my pictures were already exported as compressed JPGs and uploaded on Google Photos and my phone pictures have been backed up on Photos for years.

This might give me some frustrations from time to time in the future (when I would have liked to access/use some of the files I had there), but I really don't think it's that big of an issue.

The main point of the NAS for me was for backup. My current computer is still UP, so I haven't lost the documents that are here.

I think I only really lost the pictures I've taken since January (pictures I hadn't treated/exported to Google Photos), so it's only some personal pictures that I took for fun. They might still be on my SD card, or recoverable from there, I'll check.

Don't worry, I shared my feeling yesterday as I was still in "shock", but I think it's not that bad :)

I'll make sure to have an external backup (online) for some critical data, as, if my house burns, I'm not covered, but I was already ready to loose my pictures in that case. So it's OK :)

I realized with the multiple ALT-F packages (never heard of them before you mentioned them) that I could use the NAS for media management, sharing and player (UPNP/dlna). This is a use case that I guess many users love and I'm actually thinking about it for someone I know who has issues with his current setup. But for me, the most important part was have my pictures secured in redundancy, and I failed it with a mis manipulation.

The error is all mine and again, I thank you deeply for all the work on ALT-F and on helping me (and many others as I've seen in the different posts on the forum).

Don't worry, everything will be fine :)