You are not logged in.

#1 2012-06-26 17:03:57

max.bra
Member
From: Bologna - Italy
Registered: 2008-06-02
Posts: 93

[kernel/systemd] systemd-readahead-collect fail to start

hi all archers,
yesterday, after a few months off, i decided to re-give a try to systemd readahead.

surprise!! time ago does not gave any benefit, now will not even start!
the service stop with this log message:

systemd-readahead-collect[xxx]: Failed to read event: Value too large for defined data type

this, the service status:

systemd-readahead-collect.service - Collect Read-Ahead Data
	  Loaded: loaded (/usr/lib/systemd/system/systemd-readahead-collect.service; enabled)
	  Active: active (exited) (Result: exit-code) since Tue, 26 Jun 2012 18:08:24 +0200; 42min ago
	 Process: 158 ExecStart=/usr/lib/systemd/systemd-readahead-collect (code=exited, status=1/FAILURE)
	  Status: "Collecting readahead data"
	  CGroup: name=systemd:/system/systemd-readahead-collect.service

googling i've seen many of this errors but always related to different software/packages
the older ones tipically involved 32bit and missing LARGEFILE* libraries/functions
the most recent seems to be all on x86_64 systems and EOVERFLOW buffer overflow protection kernel bug or compilation related bug.

my system is a fully updated (26/06/2012) ArchLinux x68_64
standard core repo kernel
full systemd inited, sysvinit replaced/removed
intel hw (asus p5q-e with a q9550)
i personally compiled systemd with no luck

anyone with a similar config and error?

Offline

#2 2012-06-27 08:44:31

max.bra
Member
From: Bologna - Italy
Registered: 2008-06-02
Posts: 93

Re: [kernel/systemd] systemd-readahead-collect fail to start

today it works...

yesterday my /.readahead file was that:

x86_64-unknown-linux-gnu;VERSION=2
R

systemd-readahead-collect immediately exits at first read event

today:

x86_64-unknown-linux-gnu;VERSION=2
R/var/cache/samba/brlock.tdb
/var/cache/samba/connections.tdb
/var/cache/samba/gencache_notrans.tdb
/var/cache/samba/locking.tdb
/var/cache/samba/notify.tdb
/var/cache/samba/notify_onelevel.tdb
/var/cache/samba/printer_list.tdb
/var/cache/samba/serverid.tdb
/var/cache/samba/sessionid.tdb
...

is it likely that the reading order at boot is the same as yesterday?
if yes,
  my old (yes yesterday i've, as usual, browsed samba shares) brlock.tdb gave the EOVERFLOW reading error. actual one no! was it to big?
  naaa, it's modern code largefile awared and I doubt the file could even reach the gigabyte
if no,
  i'll never know

ideas?

p.s.: why read samba cache first to boot???? hmm

Offline

#3 2012-06-27 08:59:56

max.bra
Member
From: Bologna - Italy
Registered: 2008-06-02
Posts: 93

Re: [kernel/systemd] systemd-readahead-collect fail to start

forget it
error is back again

Offline

#4 2012-11-02 16:19:29

ewaves
Member
Registered: 2012-09-01
Posts: 39

Re: [kernel/systemd] systemd-readahead-collect fail to start

same error here, did you find a solution in the meantime?

Offline

#5 2012-11-02 16:25:40

PLANTROON
Member
From: Slovakia
Registered: 2012-08-08
Posts: 61

Re: [kernel/systemd] systemd-readahead-collect fail to start

I am getting similar error.

 [(a number increasing during the boot process)]Failed to read event: Value too large for defined data type 

Offline

#6 2012-11-02 23:09:50

max.bra
Member
From: Bologna - Italy
Registered: 2008-06-02
Posts: 93

Re: [kernel/systemd] systemd-readahead-collect fail to start

no solution here
I still have regularly that log error

Offline

#7 2013-01-19 15:21:36

kelloco2
Member
Registered: 2012-02-13
Posts: 124

Re: [kernel/systemd] systemd-readahead-collect fail to start

Do you use a swap file? I have the same problem and I think it's because. hash swapfile entry in the /etc/fstab temporarily solved the problem.
https://bugs.archlinux.org/task/33459
Do people with swapfile could confirm bug?

systemctl enable systemd-readahead-collect systemd-readahead-replay

and looked at

dmesg | grep systemd-readahead

if the bug "systemd-readahead[155]: Failed to read event: Value too large for defined data type" occurs?

then hash swapfile entry in the /etc/fstab, reboot and and again look at the dmesg output.

systemctl disable systemd-readahead-collect systemd-readahead-replay

disables service and get an earlier state. Sorry for my english

PS possible that the swap file larger than 2GB

Last edited by kelloco2 (2013-01-19 15:33:04)


sorry for my english. {Arch Linux, Debian} User

Offline

#8 2013-01-19 16:45:03

max.bra
Member
From: Bologna - Italy
Registered: 2008-06-02
Posts: 93

Re: [kernel/systemd] systemd-readahead-collect fail to start

@kelloco2 - yes!! i use a swap file and it's >2Gib. maybe this is the right way but...

1) it's also true, for me, that not always my readahead service fails. seems to be nearly random. 7-8/10 times it fails.
2) /.readhaed never contains paths to swapfile. why readahead collect needs to cache swapfile?

Offline

#9 2013-01-29 14:13:57

kelloco2
Member
Registered: 2012-02-13
Posts: 124

Re: [kernel/systemd] systemd-readahead-collect fail to start

I do not know exactly why this is so. I reported upsteram bug and I report bug to systemd's bug reports and i hope they repair it.. I do not want to give up swapfile. swap partition temporary solved the problem.


sorry for my english. {Arch Linux, Debian} User

Offline

#10 2013-08-13 10:01:57

fabrydrunk
Member
From: Catanzaro, Italy
Registered: 2012-04-16
Posts: 25

Re: [kernel/systemd] systemd-readahead-collect fail to start

I have the same problem and I have a 2GB swapfile. Does anybody have any news about this bug?

Offline

#11 2013-08-13 10:21:14

fabrydrunk
Member
From: Catanzaro, Italy
Registered: 2012-04-16
Posts: 25

Re: [kernel/systemd] systemd-readahead-collect fail to start

Hey guys. I solved the bug decreasing file dimension down to 1.5GB. Now it seems to work.

Offline

Board footer

Powered by FluxBB