Skip to main content

On servers and racks and stuff...

For goodness sake order a DVD-ROM in the least. We have several 2950s and 2850s that have CD-ROM drives and are utterly useless for ESX installs. I'm now praying the DRAC card's ISO mount works. If not, big trouble!!

Also, a few other points off the top of my head:
  1. Order less DIMMs, but higher capacity, even if you have no plans to expand. It will happen, and you will throw out a ton of 512MB/1GB/2GB sticks of FB-DIMMs that nobody else wants for this very reason.
  2. Order a ILO/DRAC/etc card. Just do it. It will save your bottom later.
  3. Only use an x64 OS if you're doing a native install, otherwise use ESX or something similar. Just doesn't make sense to have single-purpose servers in this day and age unless you're talking about an absolute monster of a SQL server or something similar.
  4. Strongly consider an expansion NIC card (dual port at least) if you do not have Intel NICs on-board. Just a good redundancy strategy.
  5. Don't skimp on PDUs in your server rack - get a good quality network accessible zero-U PDU that has plenty of outlets. 1U servers with dual-PSUs can add up to a lot of needed outlets.
  6. For goodness sake use a cable colour scheme. We just started doing this and it is fantastic. We use purple for POE/phone, blue for data, white for switch/router interconnect, orange for iSCSI, and red for internet-facing cables. Yellow would be DMZ if we had any.
  7. for CAT6/power cables. Buy extras as some come DOA. The price can't be beat.
  8. Plan ahead with your cabling methods. Do a dry run if you have the opportunity - it could change your mind for the actual installation.
  9. 1 foot patch cables - just try it. Patch panel rack looks like this: 2U 48 port patch panel, 2U horizontal cable management, 2U (1 48 port GbE switch, with room for a second). Alternatively you can do 2U patch, 2U mgmt, 2x48prt switch, 2U mgmt, 2U patch, pattern repeats. See for inspiration. It looks fantastic, leaves your vertical management spaces nice and clean, and makes tracing cables super easy. It also means you don't need to label every cable, and makes the longer cables clearly defined in the rack.
  10. Honest to goodness, have a wiring diagram of the building posted in the server room. It should label every ethernet port, power outlet, and cable trunks. Relevant power information (max circuit sizes, UPS capacity, etc.) should also be included.
  11. Power drops should be labeled and load monitored for each circuit. Also, if you have a dual PSU system, each PSU should go to a separate PDU which in turn are on separate circuits, and if you're able, separate UPS and power sources. But for the average admin, separate PDUs on separate circuits is standard. Keep a tally of what's on what PDU to get an idea of load on each circuit.
  12. No more than 80% load on a circuit.
  13. Switches - when you get to 80% used ports, buy another one already. We noticed that we were using a lot of iSCSI ports. What we didn't think about was how many ports each server really needed. Dual quad-port NICs on each server really starts to add up.
  14. Network monitoring - have some mechanism for checking network stability. We recently had some DHCP servers freak out on us with massive broadcast storms of DHCP server packets. It took down the VOIP phones really fast, and took hours to isolate.
  15. VOIP phones - give them a separate VLAN in the least. Just keep the packets isolated from the data network as best possible.
  16. Velcro - buy it in large rolls. I found some in .75" x 25 yard lengths, UL rated, for around $20CAD each.
  17. Cable labeling - when needed, do it right. Get self-laminating labels, be it laser printed, or a handheld (just got a Brady Idxpert, and it's very nice). Label the cable with two pieces of data, the port, and a brief description of the function. For example, our fax line cable label says: V2-047 FAX-3010 (the pipe denotes a new line). FAX-3010 tells me it's a fax line, and for the number ending in 3010. V2-047 indicates it's a voice port, 2nd floor, port 47. The port label system was already there from the company we used to run the cables. Again, note that only long runs of cable need labels if you use the 1-foot patch panel cable method.
  18. VMware VCenter. It may look like a nice way to save a few grand, but going with multiple Essentials packs (limited to 3 hosts per VC) versus getting one VC license is NOT the way to go. Extremely annoying, and hard to manage when moving VMs around.
  19. Think long and hard about using ESXi in the corporate environment. It lacks a lot of features that make simple things like changing datastores turn into acts of glacier-like time.
  20. Document any and all procedures into a knowledge base...even though you think it won't be needed again, it will.
  21. When cabling, do a neat, professional job - show some pride in your work! Label appropriately, and with the right types of labels. Nothing looks more awkward than labels peeling off. If possible, comb bundles of cables and velcro every 18" or so. Only use zip ties for permanent spots like trunks attaching to vertical management, even then, try to avoid them.


Popular posts from this blog

DFSR - eventid 4312 - replication just won't work

This warning isn't documented that well on the googles, so here's some google fodder:

You are trying to set up replication for a DFS folder (no existing replication)Source server is 2008R2, 'branch office' server is 2012R2 (I'm moving all our infra to 2012R2)You have no issues getting replication configuredYou see the DFSR folders get created on the other end, but nothing stagesFinally you get EventID 4312:
The DFS Replication service failed to get folder information when walking the file system on a journal wrap or loss recovery due to repeated sharing violations encountered on a folder. The service cannot replicate the folder and files in that folder until the sharing violation is resolved.  Additional Information:  Folder: F:\Users$\\Desktop\Random Folder Name\  Replicated Folder Root: F:\Users$  File ID: {00000000-0000-0000-0000-000000000000}-v0  Replicated Folder Name: Users  Replicated Folder ID: 33F0449D-5E67-4DA1-99AC-681B5BACC7E5  Replication Group…

Fixing duplicate SPNs (service principal name)

This is a pretty handy thing to know:

SPNs are used when a specific service/daemon uses Kerberos to authenticate against AD. They map a specific service, port, and object together with this convention: class/host:port/name

If you use a computer object to auth (such as local service):

If you use a user object to auth (such as a service account, or admin account):

Why do we care about duplicate SPNs? If you have two entries trying to auth using the same Kerberos ticket (I think that's right...), they will conflict, and cause errors and service failures.

To check for duplicate SPNs:
The command "setspn.exe -X

C:\Windows\system32>setspn -X
Processing entry 7
MSSQLSvc/ is registered on these accounts:
CN=SQL Admin,OU=service accounts,OU=resources,DC=company,DC=local

found 1 groups of duplicate SPNs. (truncated/sanitized)

Note that y…

Logstash to Nagios - alerting based on Windows Event ID

This took way longer than it should have to get here's a config and brain dump...

You want to have a central place to analyze Windows Event/IIS/local application logs, alert off specific events, alert off specific situations.  You don't have the budget for a boxed solution.  You want pretty graphs.  You don't particularly care about individual server states.  (see rationale below - although you certainly have all the tools here to care, I haven't provided that configuration)

ELK stack, OMD, NXlog agent, and Rsyslog.  The premise here is as follows:

Event generated on server into EventLogNXlog ships to Logstash inputLogstash filter adds fields and tags to specified eventsLogstash output sends to a passive Nagios service via the Nagios NSCA outputThe passive service on Nagios (Check_MK c/o OMD) does its thing w. alerting
Open Monitoring Distribution, but the real point here is Check_MK (IIRC Icinga uses this...).  It makes Nagios easy to use and main…