Hi,

i have just upgraded my linode from 64 up to 128mb ram. Right after the update, the system started to power off suddenly without any reason. If i issue a reboot, either the system doesn't reboot at all or switch off after few minutes of activity. I've asked to support@linode.com, but as usual these guys don't care. Anyone experienced the same problem? Any suggestion?

Thanks

walter

Here you find syslog output of the last boot process. That are a lot of things that i don't know, so i hope it can help somebody more knowledgeable than me.

Jan 5 19:12:08 blackout syslogd 1.4.1#16: restart.
Jan 5 19:12:08 blackout kernel: klogd 1.4.1#16, log source = /proc/kmsg started.
Jan 5 19:12:09 blackout kernel: Cannot find map file.
Jan 5 19:12:09 blackout kernel: No module symbols loaded - kernel modules not enabled.
Jan 5 19:12:09 blackout kernel: Checking for /dev/anon on the host...Not available (open failed with errno 6)
Jan 5 19:12:09 blackout kernel: Linux version 2.4.28-linode37-1um (root@nova1.theshore.net) (gcc version 3.3.3 20040412 (Red Hat Linux 3.3.3-7)) #2 Mon Nov 22 10:08:46 EST 2004
Jan 5 19:12:09 blackout kernel: On node 0 totalpages: 32768
Jan 5 19:12:09 blackout kernel: zone(0): 32768 pages.
Jan 5 19:12:09 blackout kernel: zone(1): 0 pages.
Jan 5 19:12:09 blackout kernel: zone(2): 0 pages.
Jan 5 19:12:09 blackout kernel: Kernel command line: mem=128M fake_ide fakehd con=null con0=fd:0,fd:1 devfs=nomount root=/dev/ubda ubda=/linodes/blackout/11744.fs ubdb=/linodes/blackout/11745.fs eth0=tuntap,blackout_0,fe:fd:43:12:5c:8c token_max=400000 token_refill=512
Jan 5 19:12:09 blackout kernel: fakehd : Changing ubd_gendisk.major_name to "hd".
Jan 5 19:12:09 blackout kernel: Calibrating delay loop... 3774.87 BogoMIPS
Jan 5 19:12:09 blackout kernel: Memory: 123480k available
Jan 5 19:12:09 blackout kernel: Dentry cache hash table entries: 16384 (order: 5, 131072 bytes)
Jan 5 19:12:09 blackout kernel: Inode cache hash table entries: 8192 (order: 4, 65536 bytes)
Jan 5 19:12:09 blackout kernel: Mount cache hash table entries: 512 (order: 0, 4096 bytes)
Jan 5 19:12:09 blackout kernel: Buffer cache hash table entries: 4096 (order: 2, 16384 bytes)
Jan 5 19:12:09 blackout kernel: Page-cache hash table entries: 32768 (order: 5, 131072 bytes)
Jan 5 19:12:09 blackout kernel: Checking for host processor cmov support...Yes
Jan 5 19:12:09 blackout kernel: Checking for host processor xmm support...No
Jan 5 19:12:09 blackout kernel: Checking that ptrace can change system call numbers...OK
Jan 5 19:12:09 blackout kernel: Checking syscall emulation patch for ptrace...OK
Jan 5 19:12:09 blackout kernel: Checking that host ptys support output SIGIO...Yes
Jan 5 19:12:09 blackout kernel: Checking that host ptys support SIGIO on close...No, enabling workaround
Jan 5 19:12:09 blackout kernel: POSIX conformance testing by UNIFIX
Jan 5 19:12:09 blackout kernel: Linux NET4.0 for Linux 2.4
Jan 5 19:12:09 blackout kernel: Based upon Swansea University Computer Society NET3.039
Jan 5 19:12:09 blackout kernel: Initializing RT netlink socket
Jan 5 19:12:09 blackout kernel: Starting kswapd
Jan 5 19:12:09 blackout kernel: VFS: Disk quotas vdquot_6.5.1
Jan 5 19:12:09 blackout kernel: Journalled Block Device driver loaded
Jan 5 19:12:09 blackout kernel: Coda Kernel/Venus communications, v5.3.18, coda@cs.cmu.edu
Jan 5 19:12:09 blackout kernel: devfs: v1.12c (20020818) Richard Gooch (rgooch@atnf.csiro.au)
Jan 5 19:12:09 blackout kernel: devfs: boot_options: 0x0
Jan 5 19:12:09 blackout kernel: Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
Jan 5 19:12:09 blackout kernel: SGI XFS with no debug enabled
Jan 5 19:12:09 blackout kernel: SGI XFS Quota Management subsystem
Jan 5 19:12:09 blackout kernel: pty: 256 Unix98 ptys configured
Jan 5 19:12:09 blackout kernel: SLIP: version 0.8.4-NET3.019-NEWTTY (dynamic channels, max=256).
Jan 5 19:12:09 blackout kernel: CSLIP: code copyright 1989 Regents of the University of California.
Jan 5 19:12:09 blackout kernel: RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
Jan 5 19:12:09 blackout kernel: loop: loaded (max 8 devices)
Jan 5 19:12:09 blackout kernel: PPP generic driver version 2.4.2
Jan 5 19:12:09 blackout kernel: PPP Deflate Compression module registered
Jan 5 19:12:09 blackout kernel: PPP BSD Compression module registered
Jan 5 19:12:09 blackout kernel: Universal TUN/TAP device driver 1.5 (C)1999-2002 Maxim Krasnyansky
Jan 5 19:12:09 blackout kernel: md: linear personality registered as nr 1
Jan 5 19:12:09 blackout kernel: md: raid0 personality registered as nr 2
Jan 5 19:12:09 blackout kernel: md: raid1 personality registered as nr 3
Jan 5 19:12:09 blackout kernel: md: raid5 personality registered as nr 4
Jan 5 19:12:09 blackout kernel: raid5: measuring checksumming speed
Jan 5 19:12:09 blackout kernel: 8regs : 6060.800 MB/sec
Jan 5 19:12:09 blackout kernel: 32regs : 2217.600 MB/sec
Jan 5 19:12:09 blackout kernel: raid5: using function: 8regs (6060.800 MB/sec)
Jan 5 19:12:09 blackout kernel: md: multipath personality registered as nr 7
Jan 5 19:12:09 blackout kernel: md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
Jan 5 19:12:09 blackout kernel: md: Autodetecting RAID arrays.
Jan 5 19:12:09 blackout kernel: md: autorun ...
Jan 5 19:12:09 blackout kernel: md: ... autorun DONE.
Jan 5 19:12:09 blackout kernel: LVM version 1.0.8(17/11/2003)
Jan 5 19:12:09 blackout kernel: Initializing Cryptographic API
Jan 5 19:12:09 blackout kernel: Initializing software serial port version 1
Jan 5 19:12:09 blackout kernel: mconsole (version 2) initialized on /linodes/blackout/.uml/blackout/mconsole
Jan 5 19:12:09 blackout kernel: Partition check:
Jan 5 19:12:09 blackout kernel: hda: unknown partition table
Jan 5 19:12:09 blackout kernel: hdb: unknown partition table
Jan 5 19:12:09 blackout kernel: Initializing stdio console driver
Jan 5 19:12:09 blackout kernel: Netdevice 0 (fe:fd:43:12:5c:8c) : TUN/TAP backend -
Jan 5 19:12:09 blackout kernel: IPv6 v0.8 for NET4.0
Jan 5 19:12:09 blackout kernel: IPv6 over IPv4 tunneling driver
Jan 5 19:12:09 blackout kernel: NET4: Linux TCP/IP 1.0 for NET4.0
Jan 5 19:12:09 blackout kernel: IP Protocols: ICMP, UDP, TCP, IGMP
Jan 5 19:12:09 blackout kernel: IP: routing cache hash table of 512 buckets, 4Kbytes
Jan 5 19:12:09 blackout kernel: TCP: Hash tables configured (established 8192 bind 16384)
Jan 5 19:12:09 blackout kernel: IPv4 over IPv4 tunneling driver
Jan 5 19:12:09 blackout kernel: GRE over IPv4 tunneling driver
Jan 5 19:12:09 blackout kernel: Linux IP multicast router 0.06 plus PIM-SM
Jan 5 19:12:09 blackout kernel: ip_conntrack version 2.1 (964 buckets, 7712 max) - 288 bytes per conntrack
Jan 5 19:12:09 blackout kernel: ip_tables: (C) 2000-2002 Netfilter core team
Jan 5 19:12:09 blackout kernel: ipt_recent v0.3.1: Stephen Frost <sfrost@snowman.net>. http://snowman.net/projects/ipt_recent/
Jan 5 19:12:09 blackout kernel: arp_tables: (C) 2002 David S. Miller
Jan 5 19:12:09 blackout kernel: NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
Jan 5 19:12:09 blackout kernel: ip6_tables: (C) 2000-2002 Netfilter core team
Jan 5 19:12:09 blackout kernel: registering ipv6 mark target
Jan 5 19:12:09 blackout kernel: NET4: Ethernet Bridge 008 for NET4.0
Jan 5 19:12:09 blackout kernel: 802.1Q VLAN Support v1.8 Ben Greear <greearb@candelatech.com>
Jan 5 19:12:09 blackout kernel: All bugs added by David S. Miller <davem@redhat.com>
Jan 5 19:12:09 blackout kernel: EXT3-fs: INFO: recovery required on readonly filesystem.
Jan 5 19:12:09 blackout kernel: EXT3-fs: write access will be enabled during recovery.
Jan 5 19:12:09 blackout kernel: kjournald starting. Commit interval 5 seconds
Jan 5 19:12:09 blackout kernel: EXT3-fs: recovery complete.
Jan 5 19:12:09 blackout kernel: EXT3-fs: mounted filesystem with ordered data mode.
Jan 5 19:12:09 blackout kernel: VFS: Mounted root (ext3 filesystem) readonly.
Jan 5 19:12:09 blackout kernel: Adding Swap: 263160k swap-space (priority -1)
Jan 5 19:12:09 blackout kernel: EXT3 FS 2.4-0.9.19, 19 August 2002 on ubd(98,0), internal journal
Jan 5 19:12:13 blackout kernel: eth0: no IPv6 routers present

I'm actually having the EXACT same problem with my linode on host38, it'll boot fine, then just randomly power off. I've looked at the logs on NOTHING specific appears right before the power off. Sometimes the logs will just have the console login screen displayed. Neither is there any rhym or reason as to when it'll shut off. This just recently started about a half hour ago.

Oh, and I literally haven't changed anything in my linode to cause this to happen, and it's been running fine for the last couple of months.

Anybody have any suggestions?

Ok, is this not enough for these linode guys to give a look at this machine?

I thought it was something related to the ram upgrade process, but now i guess it's an hardware problem.

bgmncwj: how long does it take your host to boot? Mine takes just 2 seconds (or at least this is reported in the job web page). I believe the boot process doesn't starts at all.


walter

According to the website it takes around 2-3 seconds, though it does take longer than that. I think the website just reports how long the process is in the queue, ie. only till power on.

I can watch my system boot by sshing into the host machine, and sometimes (other times it'll just turn off during boot) it'll boot just fine and give me a logon and everything. But at a random point beyond that (under 5 minutes normally) it'll just "power off" again, with no trace in any of the logs.

If this is the same problem I'm having, caker has been working on it since November, trying to figure it out. It has to do with the host kernel (not the linode kernel) near as I can tell. 2.6.8.1 or higher on the host and the problem starts happening. The older host kernels (2.6.4 if I remember?) were fine. 2.6 or 2.4 in the linode, both are affected.

I was wondering if more people were seeing the problem, I had a feeling they were but there wasn't much talk on the forums of it.

Check your /proc/cpuinfo and see what version is running on your host, see if it matches what I described above.

I have two linodes, both are suffering from this same problem now that their hosts are at 2.6.8.1 or higher. One only crashes every week or more, the other crashes every 1 to 4 days. Some pattern of activity on the host must be triggering it. I don't know what though.

Check out this thread:

http://www.linode.com/forums/viewtopic.php?t=1354

Same problem, but I also posted a small script in there that you can run from a remote host (your linux box at home for example) that will check your linode every 15 minutes, and boot it if it is down.

Edit: walter - you say you upgraded from 64 meg to 128 meg. Did you do that by just buying an instant memory upgrade, or did you migrate to a new host, going from a linode 64 to a linode 128?

?? Next time you encounter problems, please use the support ticket system.

I'm taking care of this right now, seconds after someone else posted a ticket.

-Chris

Ok. I'll do that. Thanks.

I have just bought instant memory from the web management system. Then i issued a reboot. After that, i've started to experience the problem. Now the system seems to be stable. Probably caker is working on it.

walter

Yeah, I think he fixed it, according to http://www.linode.com/forums/viewtopic.php?t=1382 .