It is impossible to log into a container after repair mode was initiated through Parallels Virtual Automation Power Panel:
Management > Maintenance > Repair Mode.
SSH attempt returns following error:
ssh_exchange_identification: Connection closed by remote host
Fail counters (last column) are present in temporary container's (ID 10000000) user_beancounters file:
[root@pvcfl~]# awk '$6>0' /proc/bc/10000000/resources numothersock 80 80 80 80 684
Normally last column should be equal to 0 - it would mean that usage of this resource is not reaching the limit. If the amount of failcounters is >0, that would mean the container is lacking this resource, and container's memory limits should be reconsidered.
numtcpsock and numothersock limits are insufficient, they are defined in /var/opt/pva/agent/etc/repair/repairve.conf:
[root@pvcfl ~]# grep SOCK= /var/opt/pva/agent/etc/repair/repairve.conf NUMTCPSOCK="80:80" NUMOTHERSOCK="80:80"
~# pva-setup --install
If you are not planning to update PVA Agent right now, or if your host is not compatible with version 6.0 - you may increase these limits up to 400 in the configuration file:
[root@pvcfl ~]# grep SOCK= /var/opt/pva/agent/etc/repair/repairve.conf NUMTCPSOCK="400:400" NUMOTHERSOCK="400:400"