Archived Forum

Sorry to hear your having trouble, try posting your logs here maybe we can help debug. Also try scanning the hdd for errors / bad blocks etc 

I'm having similar problems except in my case the Pi appears to have rebooted (looking at the logs) but won't come back up fully until I power cycle the HDD (has it's own supply and is running via a usb to sata adaptor), sometimes I have to reboot a second time. It seems to coincide with high load (I'm also using the the HDD for backups via samba) so I'm starting to suspect psu issues. The system appears to be more stable at lower overclocks so if you're overclocking it might be worth dropping down a peg or two. I've invested in a Pi Hub from Pimoroni and have just switched over to this to see if that helps (has a dedicated high power usb socket to power the Pi).

If instability persists I'm thinking of implementing a watchdog and/or monitoring for reboots and power cycling the HDD automatically.

Failing all that I'll probably revert back to my previously rock solid SD Card system and maybe set up to store and forward to a PC based installation of Emoncms.

mate could be the pi

i have got 2 dead ones in draw

the symptoms were reboots and lockups  i tried everything then one by one just died

just saying

cheers

Lasted 4 days then locked up (wouldn't even respond to a ping) rebooted and 20 mins later locked up again. However on the second occasion I'd left a putty session running and the console window was full of "journal commit IO error" entries so it looks like it may be a HDD or HDD controller problem. No clues that i can decipher in the syslog other than the the times recorded are some 15 mins or so before the lockup until the reboot is complete

e.g.

Jan 30 16:08:37 localhost dhclient: bound to 192.168.2.90 -- renewal in 1451 seconds.
Jan 30 16:09:01 localhost /USR/SBIN/CRON[3876]: (root) CMD (  [ -x /usr/lib/php5/maxlifetime ] && [ -d /var/lib/php5 ] && find /var/lib/php5/ -depth -mindepth 1 -maxdepth 1 -type f -ignore_readdir_race -cmin +$(/usr/lib/php5/maxlifetime) ! -execdir fuser -s {} 2>/dev/null \; -delete)
Jan 30 16:17:01 localhost /USR/SBIN/CRON[3973]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Jan 30 16:20:01 localhost /USR/SBIN/CRON[4016]: (smmsp) CMD (test -x /etc/init.d/sendmail && /usr/share/sendmail/sendmail cron-msp)
Jan 30 16:32:48 localhost dhclient: DHCPREQUEST on eth0 to 192.168.2.1 port 67
Jan 30 16:32:48 localhost dhclient: DHCPACK from 192.168.2.1
Jan 30 16:32:49 localhost dhclient: bound to 192.168.2.90 -- renewal in 1499 seconds.
Jan 30 16:39:01 localhost /USR/SBIN/CRON[4274]: (root) CMD (  [ -x /usr/lib/php5/maxlifetime ] && [ -d /var/lib/php5 ] && find /var/lib/php5/ -depth -mindepth 1 -maxdepth 1 -type f -ignore_readdir_race -cmin +$(/usr/lib/php5/maxlifetime) ! -execdir fuser -s {} 2>/dev/null \; -delete)
Jan 30 16:40:01 localhost /USR/SBIN/CRON[4294]: (smmsp) CMD (test -x /etc/init.d/sendmail && /usr/share/sendmail/sendmail cron-msp)
Jan 30 16:17:17 localhost kernel: imklog 5.8.11, log source = /proc/kmsg started.
Jan 30 16:17:17 localhost rsyslogd: [origin software="rsyslogd" swVersion="5.8.11" x-pid="2036" x-info="http://www.rsyslog.com"] start
Jan 30 16:17:17 localhost kernel: [    0.000000] Booting Linux on physical CPU 0

<etc etc. loads more bootup entries for 16:17 snipped>

Jan 30 16:17:29 localhost mysqld: 140130 16:17:29 [Note] /usr/sbin/mysqld: ready for connections.
Jan 30 16:17:29 localhost mysqld: Version: '5.5.33-0+wheezy1'  socket: '/var/run/mysqld/mysqld.sock'  port: 3306  (Debian)
Jan 30 20:47:09 localhost /etc/mysql/debian-start[2929]: Upgrading MySQL tables if necessary.
Jan 30 20:47:10 localhost /etc/mysql/debian-start[2934]: /usr/bin/mysql_upgrade: the '--basedir' option is always ignored
Jan 30 20:47:10 localhost /etc/mysql/debian-start[2934]: Looking for 'mysql' as: /usr/bin/mysql
Jan 30 20:47:10 localhost /etc/mysql/debian-start[2934]: Looking for 'mysqlcheck' as: /usr/bin/mysqlcheck
Jan 30 20:47:10 localhost /etc/mysql/debian-start[2934]: This installation of MySQL is already upgraded to 5.5.33, use --force if you still need to run mysql_upgrade
Jan 30 20:47:10 localhost /etc/mysql/debian-start[2958]: Checking for insecure root accounts.
Jan 30 20:47:10 localhost /etc/mysql/debian-start[2963]: Triggering myisam-recover for all MyISAM tables
Jan 30 20:47:13 localhost mysqld: 140130 20:47:13 [ERROR] /usr/sbin/mysqld: Table './emoncms/raspberrypi' is marked as crashed and should be repaired
Jan 30 20:47:13 localhost mysqld: 140130 20:47:13 [Warning] Checking table:   './emoncms/raspberrypi'
Jan 30 20:47:13 localhost mysqld: 140130 20:47:13 [ERROR] /usr/sbin/mysqld: Table './emoncms/rememberme' is marked as crashed and should be repaired
Jan 30 20:47:13 localhost mysqld: 140130 20:47:13 [Warning] Checking table:   './emoncms/rememberme'
Jan 30 20:47:19 localhost kernel: [   48.995183] ttyAMA0: 1 input overrun(s)

Locked up around 16:40 and rebooted at 20:47

First of all, I am no MySQL expert, but a bit of 'Googling' always helps:  Firstly I would backup your card saving it under a dated name (good practice).  Second, can you get to a shell prompt and login?  Or do you not get that far?  If you can, try running 'mysqlcheck' (Google 'mysqlcheck all tables' it should  be the top answer).  I have not tried it here, so 'let the buyer beware' so to speak.  Good luck.  shell>mysqlcheck all databases 

Mines been rock solid since I updated the boot image so that it matched the kernel version, see this thread for details http://openenergymonitor.org/emon/node/3688