1. What are the main components of a server? MOBO, CPU, Memory, PSU, Controller, Backplane, Cables-sata, power, scsii, disks, heat sinks, fans, Riser Cards, Controller Battery and DIMM, CMOS Battery 2. What is BIOS? Basic Input Output System --initializes and test the system hardware components. Loads a boot loader or an OS. 3. Describe the different RAID levels? 0-Striping-No Redunancy-1 disk 1-Mirrored, Redundancy-need at least two disks 5-Striping with parity, need at least 3 disks 10-Combo of striping and mirrors. 4 disks 4. What are the layers of the OSI model? Application, Presentation, Session, Transport, Network, Data Link, Physical 5. Where would you find error messages on a Linux host ? Var/log/messages, dmesg 6. What are the different ways you could check to see how much memory a host has? the free command, information is also listed under proc/meminfo, top 7. You have a server that shows a faulted DIMM describe to me what steps you would take to resolve the fault ? Reseat DIMMs Run Mem Test Use memory tester if one is available If Mem Test does not show what DIMM is failed then do the following Take out all the memory out except for the first stick and boot it up with one different stick at a time to make sure all sticks were good. Add the memory back one stick at a time to rule out failed DIMMs and bad slots on the mother board 8. What are the different run levels of Linux? 0-halt 1-single user 2-single user with networking 3-multi user-command line 4-user defined 5-Multi user with GUI 6-Reboot 9. What are some Linux commands that you know? 10. What's the difference between single mode and multi-mode? Single is one directional, multi does both Single mode is more expensive than multimode Single has faster speed and covers greater distances 12. What are the steps for the boot process? System goes through POST Power is initialized Bios is initialized POST tests internal and external components Disks and RAID systems are initialized System determines primary boot device System reads Master Boot Record of device Boot loader starts, Grub or LILO is loaded Boot loader determines and decompresses the default kernel image Kernel loads modules for all the hardware and initializes it Kernel starts the init process Init reads /etc/inittab to determine the default runlevel Init starts processes located in the default runlevel’s rc directory Init starts processes in rc.local, TTY sessions and system comes to a login prompt 13. What is the difference between TCP and UDP? TCP has a handshake. Cares about packet loss. TCP is more reliable than UDP UDP- hard and fast. Does not care if there is package loss. 15. What's the difference between LILO and GRUB? GRUB has a cli interface and can boot over the network. LILO cannot boot over the network With GRUB it is possible to have multiple operating systems on the same machine 16. How would you check for errors during boot? Some Messages maybe visible during post, var/log/messages, dmessg 17. Describe the steps you would take to resolve a server that won't power on? check the power button, see if host powered on check the RPDU breaker has not tripped Verify that power cables are connected Use a different power source if one is available Open the case and verify that all internal power cabling is connected properly and that other internal components are seated properly Check to see if lights on motherboard are active Replace Power supply Check the power button, replace if broken Replace motherboard 18. What's the difference between a router, switch, and hub? Hubs are the least intelligent, they will forward to anything attached to it. Least Expensive Switches are layer2, and they remember where devices are connected, ideally eliminate collisions Routers are the smartest of the three, allow you to leave the network, they use routing tables, most expensive 19. You have a server that loses all data drives, Describe the steps to trouble shoot and repair the host? Verify if host can be taken down with owner Reboot the host, check if drives come back online Reseat the drives, reseat the controller and backplane, verify cabling is properly connected Check the controller to see if raid is still present Replace controller Replace back plane Replace disk cabling Customer Obsession, Bias for Action S: A bunch of console tickets cut because a console switch of an S3 rack went down T: The task is to recover the switch so automation or S3 Engineer can console into hosts A: I communicated with an L4 Tech and we think it is right to cut a sev2 to logistic for a new switch R: We deployed the new switch and console is restored for all hosts Ownership S: I worked on a sev2 on WMWPS and noticed we only have 1 psu left and no bbu in the critical spare shelf and we have a 4 hrs sla on vetting powershelf tt T: To order more parts for critical spares A: I reached out to logistic on-call and I cut a tt to logistic for more psu and bbu R: