It been a busy month with VMware Explore event and lots of new announcement at VMware Explore , due to which my post has got delayed . In todays post we will talk about the VCF upgrade from 4.3.1 to 4.4.1 NSX upgrade failed on a single node with error root partition was full on a specific host. Below the screen shot of the error which clearly tells there is space issue on the host need to free up the space.
Logged into the host and run the command du -shc* ,root was 100 % full which is not very usual we see esxi having space issues. checked the vobd.log it was reporting space issue on ramdisk. vsantraces appears to be 73% used .
To fix this issue , took the backup of vsantraced.conf file and follow below steps.
cp /etc/vmware/vsan/vsantraced.conf /etc/vmware/vsan/vsantraced.conf.back
Edit the config file and change to the following rotation settings:
VSANTRACED_ROTATE_MAX_FILES=8
VSANTRACED_ROTATE_FILE_SIZE=140
run this command, "esxcli vsan trace set -f 4 -s 200" to set Max file=4 and File size=200
Delete or move the current vsantraces files from the vsantraces folder --> cd /vsantraces/
delete this "vsantraces--*.gz"
Rebooted the Esxi host. After which the root partition was 290%
Ramdisk Size Used Available Use% Mounted on
root 32M 32M 0B 100% --
etc 28M 1M 26M 4% --
opt 32M 2M 29M 7% --
var 48M 1M 46M 2% --
tmp 256M 192K 255M 0% --
iofilters 32M 0B 32M 0% --
shm 1024M 0B 1024M 0% --
crx 1024M 0B 1024M 0% --
configstore 32M 392K 31M 1% --
configstorebkp 32M 392K 31M 1% --
vsantraces 300M 239M 60M 79% --
hostdstats 2053M 4M 2048M 0% --
nestdb 512M 10M 501M 1% --
nsx-idps 64M 10M 53M 15% --
snmptraps 1M 0B 1M 0% --
Post this retried the upgrade from SDDC and it completed successfully.
Commentaires