Only usable if you use OMV with HP Microserver Gen8 where you cannot boot from a ODD disk (connector) and inserted HDDs in RAID)
Today, a day before my holiday (murphys law). I noticed that suddenly a samba share was not responsive anymore and that OMV was running in read-only mode.
A quick scan showed first a 2,5″ HDD failure (the disk where OMV is on) and a second scan showed that the USB flash drive with grub (this starts OMV) was not accessible anymore. (probably caused due too high temperatures last week).
The RAID installation with 4 HDDs was expected to be ok.
So I grabbed a new flash drive, and a new 2,5″ 500GB disk (I had those laying around especially for this purpose). a bootable SD card with OMV 4 was also present
1. change the 2,5″ disk (remove the failed one and connect the new disk)
2. insert the SD card, boot and OMV splash screen should be shown, install OMV on the 2,5″
3. After reboot insert the USB flash drive after OMV is booted
4. my USB drive is /dev/sdb
5. clean the flash drive
dd if=/dev/zero of=/dev/sdb bs=512 count=1
6. fdisk the flash drive as a bootable linux drive:
I believe the sequence as:
(toggles boot flag)
7. format the drive: mkfs.ext3 /dev/sdb1 (please notice the 1 !!, not using this issues with grub !! (Message to self)
8. mount the drive with mount /dev/sdb1 /mnt
9. grub-install –no-floppy –root-directory=/mnt /dev/sdb
Please notice that it is used: dash dash no-floppy dash dash root-directory
10. grub-mkconfig -o /mnt/boot/grub/grub.cfg
Message to self: if there is any issue reported in the screen about ext2, or about blocks .. no correct installation and usage of grub and your system will not boot
A lot of people do report issues with the Calibre database when the database is on a NAS or network drive. Last week I had issues with Calibre-Web and the validation of the database when Calibre and Calibre-web do maintenance on the database.
People get: “CantOpenError: CantOpenError: unable to open database file”
Within the comments the git repository of Calibre-web I got a hint. and that solved my issue
a number of years I was running a 13K ebook database, but I was never satisfied. Adding books was almost impossible without syncing, copying books, move the database and messing up the database with duplicates. Terrible.
Last week the database was not able to fix itself anymore and I decided to start over with only those books I really like or wanted to keep. And I wanted to fix the database issues.
To fix it, you must be able to run a NAS or (virtual) Server with SAMBA version1 (NT1). As I am running VMWARE a new Linux Server was built quickly. Please be noticed that I do not care about security in my home network on that machine and for that purpose.
Import configuration steps in /etc/samba/smb.conf
min protocol = NT1
max protocol = NT1
client min protocol = NT1
client max protocol = NT1
# this will run samba in Samba Version 1 mode.
force user = [a-username]
path = /var/www/html/books
writeable = yes
browseable = yes
guest ok = yes
guest account = [a-username]
create mask = 0777
directory mask = 0777
oplocks = false
level2 oplocks = false
# Especially the oplocks on false are needed. If not Calibre-Web will not get access to the database if Calibre is using it. If that happens your database can be corrupted again very quickly.
If you run Calibre under Windows 10 make sure you add SMB1 support through the ‘features’ option this is disabled by default.
I started with 400 books and now I added 520 books in many sessions without any open database issue.
Using the oplocks on false could be also a solution when using SMBv3 or SMBv4 but I did not try that.
Please also notice that the writer of Calibre does not suggest to use a network share and it is known that for some it is working fine and for others not. This blog items could lead to more solution(s). Even it is not supported ;)
The long awaited feature: add manually samba shares has been implemented. Not functional yet. (it is showing issues during our initial tests)
The sample image is taken from a firmware version not available. So do not bother to ask for it, you will not get it ;_)
status update about 1.04RC5 firmware and reported bugs
1. After reboot, the time is 1 hour behind: can be reproduced and will be fixed now
2. Firefox cannot change password from user: can be reproduced and will be fixed now
3. HDD cannot be in standby: cannot be reproduced
4. Unicode part under review when mounting samba share with unicode characters (not displayed properly on cfullhdma, when using nfs no problem
Is the following line for mounting samba shares a good one?
smbmount //server/share /tmp/mountpoing -o lfs,codepage=cp437,iocharset=utf8,username=xx,password=xx
Or has someone a better mount line? Like something with CIFS?
The Conceptronic CFULLHDMA has an automatical way of finding Samba Shares, to make life easy for customers. The device is using ‘smbtree’ for this purpose. smbtree is a text based network browser.
As I do a lot of testings I suddenly did not get any samba share to see on the CFULLHDMA anymore. Bummer, a bug? let’s find out …
Normally I have 6!! Samba shared devices in my network. Suddenly, after rebooting a PC (XP) and a NAS I lost all Samba shares to bee seen in my CFULLHDMA, reboot, turn off / turn on .. nothing helped.
Analyzing I found out following:
the smbtree was generating an error on the CFULLHDMA.
session request to 192.168.1.30 failed (Called name not present)
Data overflow in cli_receive_trans
the 192.168.1.30 is my printserver, turning off that device only gave totally no output on running smbtree.
Using the same utility ‘smbtree’ on my Centos webserver (where aroundmyroom.com is running at) generated the same error. For me it meant that it was not a bug on my CFULLHDMA but a general ‘error’ within my network. Checking my network neighbourhood at my XP machine revealed only 1 samba share: a rebooted nas device I had to reset to factory defaults and modified on IP level again (fixed) and where I changed its Samba Share name as well. The rest of my shares were even not seen at my XP system, Linux Server or CFULLHDMA.
To solve the issue I did following:
I rebooted my DSL modem (router / dns / dhcp device) &