Good Day,
I've had two proxmox boxes do this...
both using v4.0-48 then apt-get upgrade for the latest deb patches etc.
If I try to start prox's console.. I see this...
Error 500: starting worker failed: unable to redirect STDIN - No such device or address at /usr/share/perl5/PVE/RPCEnvironment.pm line 840.
If I try to SSH, the connection is dropped/refused.
Which leaves me in a position of not being able to get into the box to see what's going on/shut down gracefully.
This managed to trash the filesystem on a dirty power off of the box I was testing.
loads of inode fixing on re power up, so had to rebuild prox.
(HP P400 RAID cache controller.)
On the other box with the same problem I've left running. but having the same errors as above.
All the VMs are running quite happily, but I can't control them from Prox.
If I initiate a shutdown of a VM.. again, I see this .
starting worker failed: unable to redirect STDIN - No such device or address at /usr/share/perl5/PVE/RPCEnvironment.pm line 840. (500)
I've not done anything to these installs bar this.
The change I made was to /etc/defaults/snmpd as syslog was so noisy with snmp reports.
I can't see how this has broken things.. but... stranger things have happened.
In /etc/default/snmpd, I changed
SNMPDOPTS='-Lsd -Lf /dev/null -u snmp -I -smux -p /var/run/snmpd.pid -g root 0.0.0.0'
to
-LSwd and -LS4d, which was accepted as the service restarted without errors.
....perplexed....
Any suggestions how I can get back in to undo this changes on a live box, I really don't want to just power off the box because of the experiences of last time. However, this box is just a software RAID1 machine that I was using while I worked on the main server. (DL380G5) with no caching controller like the HP.....
If I can close down gracefully, I maybe able to undo the changes via a live sys-rescue disk.... and prox will recover from whatever it's decided to do?
Any thoughts as to what may of happened or way forward?
Many Thanks.
M
I've had two proxmox boxes do this...
both using v4.0-48 then apt-get upgrade for the latest deb patches etc.
If I try to start prox's console.. I see this...
Error 500: starting worker failed: unable to redirect STDIN - No such device or address at /usr/share/perl5/PVE/RPCEnvironment.pm line 840.
If I try to SSH, the connection is dropped/refused.
Which leaves me in a position of not being able to get into the box to see what's going on/shut down gracefully.
This managed to trash the filesystem on a dirty power off of the box I was testing.
loads of inode fixing on re power up, so had to rebuild prox.
(HP P400 RAID cache controller.)
On the other box with the same problem I've left running. but having the same errors as above.
All the VMs are running quite happily, but I can't control them from Prox.
If I initiate a shutdown of a VM.. again, I see this .
starting worker failed: unable to redirect STDIN - No such device or address at /usr/share/perl5/PVE/RPCEnvironment.pm line 840. (500)
I've not done anything to these installs bar this.
The change I made was to /etc/defaults/snmpd as syslog was so noisy with snmp reports.
I can't see how this has broken things.. but... stranger things have happened.
In /etc/default/snmpd, I changed
SNMPDOPTS='-Lsd -Lf /dev/null -u snmp -I -smux -p /var/run/snmpd.pid -g root 0.0.0.0'
to
-LSwd and -LS4d, which was accepted as the service restarted without errors.
....perplexed....
Any suggestions how I can get back in to undo this changes on a live box, I really don't want to just power off the box because of the experiences of last time. However, this box is just a software RAID1 machine that I was using while I worked on the main server. (DL380G5) with no caching controller like the HP.....
If I can close down gracefully, I maybe able to undo the changes via a live sys-rescue disk.... and prox will recover from whatever it's decided to do?
Any thoughts as to what may of happened or way forward?
Many Thanks.
M