Next round of reboots for the roll out of a new kernel version:
- Hadoop worker nodes - analytics10[28-77]
- Hadoop master nodes - analytics100[12] (soon to be replaced with analytics-master100[12])
- Hadoop coordinator - analytics1003
- AQS nodes (aqs1001-1009)
- Druid Private nodes (druid1001-3)
- Druid Public nodes (druid1004-6)
- Kafka Jumbo
- Kafka main codfw
- Kafka main eqiad
- Kafka Analytics
- stat100[4-6] hosts
- notebook100[3,4]
- conf100[4-6]
- db110[7,8]
- eventlog1002
- dbstore1002 (will not be done since old Trusty, will be replaced soon)
New hosts will follow as soon as the new kernel will be deployed.
Before rebooting please do the following checks to ensure that PXE is not the preferred option (so a reboot will not trigger a reimage):
- ipmitool -I lanplus -H "HOSTNAME" -U root -E chassis bootparam get 5 | awk '{ FS=":" }; /(Boot parameter data|Boot Device Selector)/{ print $2 }' (should be all zeros - to check)
- in the mgmt console, racadm get bios.BiosBootSettings.BootSeq that should be BootSeq=HardDisk.List.1-1,NIC.Integrated.1-1-1.