Bug#1058706: Aw: Re: Bug#1058706: mariadb-server-core: mariadbd fails to start on fresh install on armel

2023-12-19 Thread Tuukka Pasanen

Hello,

I've investigate this yesterday and I could reproduce this one:

Dec 17 11:17:55 gast1 mariadbd[1291]: 2023-12-17 11:17:55 0 [ERROR] Could not open 
mysql.plugin table: "Table 'mysql.plugin' doesn't exist". Some plugins may be 
not loaded
Dec 17 11:17:55 gast1 mariadbd[1291]: 2023-12-17 11:17:55 0 [ERROR] 
Unknown/unsupported storage engine: InnoDB
Dec 17 11:17:55 gast1 mariadbd[1291]: 2023-12-17 11:17:55 0 [ERROR] Aborting

Can you do (Please on testing machine not production will destroy your 
database!!!)
apt purge mariadb-server
# check is your database is not available /var/lib/mysql if it is remote it
mkdir -p /var/lib/mysql
chown mysql:mysql /var/lib/mysql
and then re-install mariad-server again. Does it work then? Just trying to see 
if my findings are correct. What comes to this InnoDB corruption I can't 
reproduce that.

Sincerely,
Tuukka

k-...@web.de kirjoitti 19.12.2023 klo 12.18:

Hi Tuukka,
I don't think the issue is related to the armel hardware.
I see the same error on my XEN-VM (1 VCPU) based on a Intel(R) Celeron(R) CPU 
G3900 @ 2.80GHz processor with 2 cores.
No lack of memory could be seen during installation of mariadb-server package. 
I tested it with 2 or 8 GB of assigned RAM.

Regards,
Konstantin
   


Gesendet: Montag, 18. Dezember 2023 um 08:23 Uhr
Von: "Tuukka Pasanen"
An: "Scott Barker",1058...@bugs.debian.org, "Otto 
Kekäläinen",k-...@web.de
Cc: "Faustin Lammler"
Betreff: Re: Bug#1058706: mariadb-server-core: mariadbd fails to start on fresh 
install on armel
Hello,

Have you monitored that is memory exhausted at the installation? This
could lead error like that.

I assume your XEN/KVM/Docker base machine is some older system with not
so fancy CPU or is there any particular reason to use armel build?

Sincerly,

Tuukka

Scott Barker kirjoitti 17.12.2023 klo 19.15:

Same result with the latest kernel:

Linux nas-1 6.1.0-16-marvell #1 Debian 6.1.67-1 (2023-12-12) armv5tel
GNU/Linux

On Sun, Dec 17, 2023 at 09:59:23AM -0700, Scott Barker wrote:

Prior to the installation, /var/lib/mysql did not exist. After
installation:

61559 4 drwxr-xr-x 2 mysql  mysql  4096 Dec 17 09:56
/var/lib/mysql
61975 4 -rw-rw 1 mysql  mysql    52 Dec 17 09:56
/var/lib/mysql/aria_log_control
61983 12304 -rw-rw 1 mysql  mysql  12582912 Dec 17 09:56
/var/lib/mysql/ibdata1
61963 0 -rw-r--r-- 1 root   root  0 Dec 17 09:56
/var/lib/mysql/debian-10.11.flag
61990 98404 -rw-rw 1 mysql  mysql 100663296 Dec 17 09:56
/var/lib/mysql/ib_logfile101
61982    16 -rw-rw 1 mysql  mysql 16384 Dec 17 09:56
/var/lib/mysql/aria_log.0001

uname -a reports:

Linux nas-1 5.10.0-26-marvell #1 Debian 5.10.197-1 (2023-09-29)
armv5tel GNU/Linux

On Sun, Dec 17, 2023 at 11:48:16PM +0800, Otto Kekäläinen wrote:

Hi Scott and Kr!

Did you note this line?

2023-12-14 14:51:33 0 [Note] InnoDB: Header page consists of zero
bytes in
datafile: ./ibdata1, Space ID:0, Flags: 0

Can you include the output of  `find /var/lib/mysql -ls` so we can
see what
files your system has?

What is the hardware you have? The report shows 'Linux
5.10.0-26-marvell '.
What does `uname -a` yield?

The test suite passes on armhf at
e.g.https://launchpad.net/~mysql-ubuntu/
+archive/ubuntu/mariadb-10.11/+builds?build_text=_state=all.
At https://
ci.debian.net/packages/m/mariadb/unstable/armel/ it fails due to
regression in
subselect which is unrelated to what you reported.


--
Scott Barker

--
To unsubscribe, send mail to1058706-unsubscr...@bugs.debian.org.


Bug#1058706: Aw: Re: Bug#1058706: mariadb-server-core: mariadbd fails to start on fresh install on armel

2023-12-19 Thread k-i-r
Hi Tuukka,
I don't think the issue is related to the armel hardware.
I see the same error on my XEN-VM (1 VCPU) based on a Intel(R) Celeron(R) CPU 
G3900 @ 2.80GHz processor with 2 cores.
No lack of memory could be seen during installation of mariadb-server package. 
I tested it with 2 or 8 GB of assigned RAM.

Regards,
Konstantin 
  

Gesendet: Montag, 18. Dezember 2023 um 08:23 Uhr
Von: "Tuukka Pasanen" 
An: "Scott Barker" , 1058...@bugs.debian.org, "Otto 
Kekäläinen" , k-...@web.de
Cc: "Faustin Lammler" 
Betreff: Re: Bug#1058706: mariadb-server-core: mariadbd fails to start on fresh 
install on armel
Hello,

Have you monitored that is memory exhausted at the installation? This
could lead error like that.

I assume your XEN/KVM/Docker base machine is some older system with not
so fancy CPU or is there any particular reason to use armel build?

Sincerly,

Tuukka

Scott Barker kirjoitti 17.12.2023 klo 19.15:
> Same result with the latest kernel:
>
> Linux nas-1 6.1.0-16-marvell #1 Debian 6.1.67-1 (2023-12-12) armv5tel
> GNU/Linux
>
> On Sun, Dec 17, 2023 at 09:59:23AM -0700, Scott Barker wrote:
>> Prior to the installation, /var/lib/mysql did not exist. After
>> installation:
>>
>> 61559 4 drwxr-xr-x 2 mysql  mysql  4096 Dec 17 09:56
>> /var/lib/mysql
>> 61975 4 -rw-rw 1 mysql  mysql    52 Dec 17 09:56
>> /var/lib/mysql/aria_log_control
>> 61983 12304 -rw-rw 1 mysql  mysql  12582912 Dec 17 09:56
>> /var/lib/mysql/ibdata1
>> 61963 0 -rw-r--r-- 1 root   root  0 Dec 17 09:56
>> /var/lib/mysql/debian-10.11.flag
>> 61990 98404 -rw-rw 1 mysql  mysql 100663296 Dec 17 09:56
>> /var/lib/mysql/ib_logfile101
>> 61982    16 -rw-rw 1 mysql  mysql 16384 Dec 17 09:56
>> /var/lib/mysql/aria_log.0001
>>
>> uname -a reports:
>>
>> Linux nas-1 5.10.0-26-marvell #1 Debian 5.10.197-1 (2023-09-29)
>> armv5tel GNU/Linux
>>
>> On Sun, Dec 17, 2023 at 11:48:16PM +0800, Otto Kekäläinen wrote:
>>> Hi Scott and Kr!
>>>
>>> Did you note this line?
>>>
>>> 2023-12-14 14:51:33 0 [Note] InnoDB: Header page consists of zero
>>> bytes in
>>> datafile: ./ibdata1, Space ID:0, Flags: 0
>>>
>>> Can you include the output of  `find /var/lib/mysql -ls` so we can
>>> see what
>>> files your system has?
>>>
>>> What is the hardware you have? The report shows 'Linux
>>> 5.10.0-26-marvell '.
>>> What does `uname -a` yield?
>>>
>>> The test suite passes on armhf at
>>> e.g. https://launchpad.net/~mysql-ubuntu/
>>> +archive/ubuntu/mariadb-10.11/+builds?build_text=_state=all.
>>> At https://
>>> ci.debian.net/packages/m/mariadb/unstable/armel/ it fails due to
>>> regression in
>>> subselect which is unrelated to what you reported.
>>>
>>
>> --
>> Scott Barker
>>
>> --
>> To unsubscribe, send mail to 1058706-unsubscr...@bugs.debian.org.
>



Bug#1058706: Aw: Re: Bug#1058706: mariadb-server-core: mariadbd fails to start on fresh install on armel

2023-12-17 Thread k-i-r


Hi Otto,
 

root@gast1:~# find /var/lib/mysql -ls
   262301  4 drwxr-xr-x   2 mysql    mysql    4096 Dec 17 11:17 
/var/lib/mysql
   263594  4 -rw-rw   1 mysql    mysql  52 Dec 17 11:17 
/var/lib/mysql/aria_log_control
   263597  98304 -rw-rw   1 mysql    mysql    100663296 Dec 17 11:17 
/var/lib/mysql/ib_logfile101
   263596  12288 -rw-rw   1 mysql    mysql 12582912 Dec 17 11:17 
/var/lib/mysql/ibdata1
   263591  0 -rw-r--r--   1 root root 0 Dec 17 11:17 
/var/lib/mysql/debian-10.11.flag
   263595 16 -rw-rw   1 mysql    mysql    16384 Dec 17 11:17 
/var/lib/mysql/aria_log.0001
root@gast1:~#
 
root@gast1:~# uname -a
Linux gast1 6.1.0-15-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.66-1 (2023-12-09) 
x86_64 GNU/Linux
root@gast1:~#

Regards,
Konstantin

 

Gesendet: Sonntag, 17. Dezember 2023 um 16:48 Uhr
Von: "Otto Kekäläinen" 
An: "Scott Barker" , 1058...@bugs.debian.org, k-...@web.de
Cc: "Faustin Lammler" , "Tuukka Pasanen" 

Betreff: Re: Bug#1058706: mariadb-server-core: mariadbd fails to start on fresh 
install on armel

Hi Scott and Kr!
 
Did you note this line?
 
2023-12-14 14:51:33 0 [Note] InnoDB: Header page consists of zero bytes in 
datafile: ./ibdata1, Space ID:0, Flags: 0
 
Can you include the output of  `find /var/lib/mysql -ls` so we can see what 
files your system has?
 
What is the hardware you have? The report shows 'Linux 5.10.0-26-marvell '. 
What does `uname -a` yield?
 
The test suite passes on armhf at e.g. 
https://launchpad.net/~mysql-ubuntu/+archive/ubuntu/mariadb-10.11/+builds?build_text=_state=all.
 At 
https://ci.debian.net/packages/m/mariadb/unstable/armel/[https://ci.debian.net/packages/m/mariadb/unstable/armel/]
 it fails due to regression in subselect which is unrelated to what you 
reported.