M3 B1 (x86_64) boot failure after install - assertion

This forum is dedicated to testing early releases and cauldron : Howtos, tips, tricks and user global feedback and thoughts...

Helpful tip :
For bugs tracking we use : https://bugs.mageia.org = The Mageia Bug Tracker
In this bug tracker you'll find already reported bugs and you'll be able to report those you have found....

M3 B1 (x86_64) boot failure after install - assertion

Postby heymanj » Dec 27th, '12, 18:20

OK, in my attempt to stress Mageia3 Beta1 x86_64, I might have done too much initially.

Lenovo w500 laptop 4GB w/Memory
Installed to a 32GB USB memory stick (/dev/sdb)
two encrytped file systems (/ and /home), /boot unencrypted
boot loader installed to the memory stick
Wired ethernet

It took upwards of 24hrs to do the install, but that I am willing to bet was the USB memory stick - though I would have thought
it would have been faster than with a disk.

on system reboot (after the install) I got the splash screen, the prompt for my passphrase and then the following

systemd-fsck[714]: /dev/sdb1: clean, 40/73152 files, 16545/292608 blocks
[ 105.165187] systemd[1]: Assertion 'j->installed' failed at src/core/transaction.c:553, function transaction_appy(). Aborting.

By throwing the installation onto a memory stick, did I exceed what should be done to a first beta?

jerry
Software is the difference between hardware and reality
heymanj
 
Posts: 18
Joined: Jun 6th, '11, 21:52
Location: Raleigh, NC USA

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby marja » Dec 28th, '12, 17:25

heymanj wrote:OK, in my attempt to stress Mageia3 Beta1 x86_64, I might have done too much initially.

Lenovo w500 laptop 4GB w/Memory
Installed to a 32GB USB memory stick (/dev/sdb)


heymanj wrote:
By throwing the installation onto a memory stick, did I exceed what should be done to a first beta?

jerry


This is probably bug 1215 https://bugs.mageia.org/show_bug.cgi?id=1215
With installer (diskdrake) it is impossible to partition a SSD well. It is better to do the partioning with, for instance, gparted. The settings should be:
Align to: MiB
and
Free space preceding (MiB): 2
User avatar
marja
 
Posts: 546
Joined: Aug 22nd, '11, 20:50

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby doktor5000 » Dec 29th, '12, 20:11

marja wrote:
This is probably bug 1215 https://bugs.mageia.org/show_bug.cgi?id=1215
With installer (diskdrake) it is impossible to partition a SSD well

Well, as already mentioned, this should be rechecked and verified. IMHO this is not true: https://wiki.mageia.org/en/Mageia_2_Rel ... tes#Stage2
This was commited in August 2011: https://bugs.mageia.org/show_bug.cgi?id=1215#c10 as http://svnweb.mageia.org/soft?view=revi ... ision=1844

So currently partitioning via installer should be okay for any SSD/flashdrive with erase block sizes below 1MB, (128K, 265K and 512K, which most if not all SSDs are using currently)

You may want to check https://wiki.archlinux.org/index.php/Pa ... l_Overview or http://en.wikipedia.org/wiki/Flash_memory#Block_erasure for context information.
Cauldron is not for the faint of heart!
Caution: Hot, bubbling magic inside. May explode or cook your kittens!
----
Disclaimer: Beware of allergic reactions in answer to unconstructive complaint-type posts
User avatar
doktor5000
 
Posts: 18051
Joined: Jun 4th, '11, 10:10
Location: Leipzig, Germany

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby heymanj » Dec 29th, '12, 20:59

Additional information on the flash drive:

Model EKMMD32GC400
32GB
Read Speed: up to 15 MB/s
Write Speed: up to 8 MB/s

I cannot seem to find the information on the erase block size as mentioned in the previous post.

I deleted the installed FAT partition, created /boot (1GB), swap (4GB), / (14.6GB encrypted), and /home (10GB encrypted).
The USB flash drive can be mounted on my Mageia 2 system w/o any issues and I can access all filesystems.

There is a 10MB core file at / on the flash drive. I can provide that if someone wants to look at it, but due to it's size I will not just append it here without request.
Software is the difference between hardware and reality
heymanj
 
Posts: 18
Joined: Jun 6th, '11, 21:52
Location: Raleigh, NC USA

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby doktor5000 » Dec 29th, '12, 21:33

heymanj wrote:I cannot seem to find the information on the erase block size as mentioned in the previous post

You will probably not find that information anywhere, besides for most popular SSD drives, and even there not all
vendors publish those values, as those are regarded as something like drive internals ... :/

That's why people want to be on the safe side and align partitions on 2MB boundaries, as that
is a multiple of all currently known erase block sizes, you could even use 4MB or 8MB or so, but that would only waste more space.
Cauldron is not for the faint of heart!
Caution: Hot, bubbling magic inside. May explode or cook your kittens!
----
Disclaimer: Beware of allergic reactions in answer to unconstructive complaint-type posts
User avatar
doktor5000
 
Posts: 18051
Joined: Jun 4th, '11, 10:10
Location: Leipzig, Germany

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby heymanj » Dec 29th, '12, 21:55

so, repartitioning the drive with the following:

1024MB /boot
4096MB swap
10240 MB /
14436 MB /home

should work?
Software is the difference between hardware and reality
heymanj
 
Posts: 18
Joined: Jun 6th, '11, 21:52
Location: Raleigh, NC USA

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby doktor5000 » Dec 29th, '12, 22:50

Maybe you misunderstood, it's not about the partition size, but the start point / end point (boundaries) needs to be properly aligned.
See marjas post for instructions how to do it with gparted.
Cauldron is not for the faint of heart!
Caution: Hot, bubbling magic inside. May explode or cook your kittens!
----
Disclaimer: Beware of allergic reactions in answer to unconstructive complaint-type posts
User avatar
doktor5000
 
Posts: 18051
Joined: Jun 4th, '11, 10:10
Location: Leipzig, Germany

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby heymanj » Dec 29th, '12, 23:09

Thanks for the clarification. installing gparted now.
Software is the difference between hardware and reality
heymanj
 
Posts: 18
Joined: Jun 6th, '11, 21:52
Location: Raleigh, NC USA

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby doktor5000 » Dec 29th, '12, 23:15

Before you change anything, why not try to first verify if your partitions are aligned properly currently or not?
Do an
Code: Select all
fdisk -l /dev/sdX
as root on the disk in question (sdX is only a placeholder) and post the output here.
Cauldron is not for the faint of heart!
Caution: Hot, bubbling magic inside. May explode or cook your kittens!
----
Disclaimer: Beware of allergic reactions in answer to unconstructive complaint-type posts
User avatar
doktor5000
 
Posts: 18051
Joined: Jun 4th, '11, 10:10
Location: Leipzig, Germany

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby heymanj » Dec 29th, '12, 23:30

It was a throw away installation, so I've already used gparted and redone the partitions per marjas' description.

Unfortunately, there is no way to use LUKS on them (that I can tell). I've told Mageia install to use existing partitions (done with gparted) to guarantee that the partitioning work under gparted stays.
Software is the difference between hardware and reality
heymanj
 
Posts: 18
Joined: Jun 6th, '11, 21:52
Location: Raleigh, NC USA

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby marja » Dec 30th, '12, 12:21

doktor5000 wrote:
marja wrote:
This is probably bug 1215 https://bugs.mageia.org/show_bug.cgi?id=1215
With installer (diskdrake) it is impossible to partition a SSD well

Well, as already mentioned, this should be rechecked and verified. IMHO this is not true: https://wiki.mageia.org/en/Mageia_2_Rel ... tes#Stage2
This was commited in August 2011: https://bugs.mageia.org/show_bug.cgi?id=1215#c10 as http://svnweb.mageia.org/soft?view=revi ... ision=1844


Thanks for correcting me, I took the fact that the bug was still open for as "committed, but not submitted because of some other issues".
User avatar
marja
 
Posts: 546
Joined: Aug 22nd, '11, 20:50

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby doktor5000 » Dec 30th, '12, 14:17

No, it was commited and it should be fixed, at least for the original report. But see Thierrys second-to-last comment
about this possibly breaking windows partitions when resizing them (which could be possibly workarounded/fixed
by only aligning partitions which are newly created)

Anyways, my last comment still stands, this should be rechecked to be sure. Create some partitions
via mga2 installer, and list them afterwards with fdisk -l -u /dev/sdX

All beginning offsets should be divisible by 8 (or to be exact -> offset MOD 8 = 0)
that's why most partitioning tools try to align by an offset of 1 or 2MB, as i tried to explain above.

Here's another good link:
http://forums.gentoo.org/viewtopic-t-82 ... t-ssd.html
Also maybe for some context:
http://web.archive.org/web/201012242119 ... lock-size/
Also maybe http://web.archive.org/web/201012241147 ... -for-ssds/ but that's more about SSDs
Cauldron is not for the faint of heart!
Caution: Hot, bubbling magic inside. May explode or cook your kittens!
----
Disclaimer: Beware of allergic reactions in answer to unconstructive complaint-type posts
User avatar
doktor5000
 
Posts: 18051
Joined: Jun 4th, '11, 10:10
Location: Leipzig, Germany

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby djennings » Jan 15th, '13, 21:21

Here is what diskdrake produced on a clean install of Mageia 3 Beta 2 on my SSD

Code: Select all
# fdisk -l -u /dev/sda

Disk /dev/sda: 128.0 GB, 128035676160 bytes, 250069680 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0007007a

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *        2048    39166469    19582211   83  Linux
/dev/sda2        39169998   250067789   105448896    5  Extended
/dev/sda5        39170048    48387779     4608866   83  Linux
/dev/sda6        48390144    56564864     4087360+  82  Linux swap / Solaris
/dev/sda7        56567808   250067789    96749991   83  Linux


All the partitions start sectors are divisible by 8 apart from the extended partition, but that does not matter, right?
User avatar
djennings
 
Posts: 613
Joined: Jun 2nd, '11, 23:51
Location: Wokingham, UK

Re: M3 B1 (x86_64) boot failure after install - assertion

Postby doktor5000 » Jan 16th, '13, 23:19

Only the primary partitions/logical drives are important, as the extended partition is only a container.
diskdrake aligns partitions at 1 MiB boundaries. But depending on the SSD this may not be enough,
as some newer SSDs use erase block sizes > 1 MiB, see http://www.ocztechnologyforum.com/forum ... s-Vertex-3 for example.
Cauldron is not for the faint of heart!
Caution: Hot, bubbling magic inside. May explode or cook your kittens!
----
Disclaimer: Beware of allergic reactions in answer to unconstructive complaint-type posts
User avatar
doktor5000
 
Posts: 18051
Joined: Jun 4th, '11, 10:10
Location: Leipzig, Germany


Return to Testing : Alpha, Beta, RC and Cauldron

Who is online

Users browsing this forum: Google [Bot] and 1 guest

cron