what did you learn today? (part 2)

[SomeHandleThatStartsWithS]

Entegy[/url]":x3lx1hrf]Canada's only full MS Store is in Toronto too.

Separated from an Apple store by a single empty store front, and a few steps from Tesla's store front.

 

[Black_Obsidian]

Alfrik[/url]":2l5v2uux]

Entegy[/url]":2l5v2uux]Canada's only full MS Store is in Toronto too.

Separated from an Apple store by a single empty store front, and a few steps from Tesla's store front.

This sounds like a place I need to visit. Where is it?

 

[Rick25]

http://www.google.ca/url?sa=t&rct=j&q=y ... 5608,d.aWM

Yorkdale Mall - it's right off the 401 at Dufferin.

 

[ncrand]

As much as I hate to admit it, Toronto IS the centre of the universe when it comes to Canada

 

[Rick25]

That Windows 8.1 still won't let you shutdown while you're logged in. I had to log out, then go to the login page to get the Power button

Still no real start menu, but usability wise it seems a little more friendly. Still think I'll wait for 9 though

 

[Barmaglot]

Huh? Right click the start button and the shut down menu is right there. Not to mention the settings menu in the charms bar. I shut down my Windows 8 machine at least twice a day without ever logging out.

 

[hawkbox]

Yeah I'm confused. Shutting down is easy while logged in. Just not in the most obvious spot.

 

[dlp]

Rick25[/url]":3a7nyz9y]That Windows 8.1 still won't let you shutdown while you're logged in. I had to log out, then go to the login page to get the Power button

Still no real start menu, but usability wise it seems a little more friendly. Still think I'll wait for 9 though

Windows 8 will, and as others said 8.1 makes it better. Go to the desktop, Alt+F4 and you get the old menu with the drop down for doing something to your session (restart, shutdown, log off, etc.). Not the nicest way to do it, but it's there.

 

[ferzerp]

windows+r "shutdown [flags]"


This is the fastest way anyway

 

[Rick25]

Thanks all, I was right clicking everywhere but never the Start button. Thanks for the info.

 

[Black_Obsidian]

Rick25[/url]":2elzckud]http://www.google.ca/url?sa=t&rct=j&q=yorkdale%20mall&source=web&cd=1&cad=rja&sqi=2&ved=0CCsQFjAA&url=http%3A%2F%2Fwww.yorkdale.com%2F&ei=jLjWUeewN4KorAHH2oD4CQ&usg=AFQjCNGB_SCqk01d42nL3kIghpjTxw12Zg&bvm=bv.48705608,d.aWM

Yorkdale Mall - it's right off the 401 at Dufferin.

Figures it'd be Yorkdale. That would have been my first guess, except I was there just a few months ago and I figured I would have noticed a Tesla storefront. Either it hadn't opened yet or shame on me...

 

[sporkme]

That a McAfee "trust product" requiring a registration to request a site be reviewed for their "Site Advisor" service sends registration emails that do not come from a valid hostname:

NOQUEUE: reject: RCPT from dnvwsnorelay2.mcafee.com[161.69.31.169]: 
450 4.1.8 <vbullet@dalsavbull1.prod.mcafee.com>: Sender address rejected: Domain not found; 
from=<vbullet@dalsavbull1.prod.mcafee.com> to=<manage@example.org> 
proto=ESMTP helo=<dnvwsnorelay2.mcafee.com>

Nice to run a service that's purporting to be on the lookout for "suspicious websites" and then send email with a made-up envelope-from address. Assclowns.

 

[FoamingPipeSnake]

I saw a post earlier about a box with 64GB of RAM, and and a pagefile to match!

I'm in a similar situation that I have a department SQL box with 64GB and the pagefile is also set to 64GB. I have no reason to reduce that, since I'm not hindered by disk space, but I wonder if I should drop that to 8GB. The only time I see the pagefile being hit is when I am re-syncing a transactional replication. Any thoughts on how I should do this?

 

[theevilsharpie]

FoamingPipeSnake[/url]":2cbw7pe7]I saw a post earlier about a box with 64GB of RAM, and and a pagefile to match!

I'm in a similar situation that I have a department SQL box with 64GB and the pagefile is also set to 64GB. I have no reason to reduce that, since I'm not hindered by disk space, but I wonder if I should drop that to 8GB. The only time I see the pagefile being hit is when I am re-syncing a transactional replication. Any thoughts on how I should do this?

SQL Server allows you to control how much RAM it uses, so there's no reason to have such a large page file on a SQL server unless you need the capability of dumping the entire contents of RAM to disk when the server BSODs.

A 6-8GB page file that allow parts of the OS, non-critical SQL components, and other utilities to be moved out of RAM is plenty. Anything more than that is a waste of disk space.

 

[kennedye]

Oracle still tends to complain if installing it on a Linux host with less than at least 1/2 physical RAM available as swap, or at least it did the last time I had to work with the DBAs to build a new box.

 

[PaveHawk-]

The day I need Enterprise Vault to be working, it isnt.

I need VMware 3.5 Enterprise license files that exist in the damn store. I have to migrate a client from 3.5 to 5.x with new physical hosts. I wanted to setup and use svMotion or even vMotion so I could get the job done quicker.

All the VMware .LIC files are in my EV archive.

The physical host that has the closed data partitions had a RAID controller + disk failure. I cant get to my damn EV archive

 

[jshiplett]

You can get a 3.5 license through the 60-day evals available on VMware Partner Central if you have access

PM me if you don't

 

[PaveHawk-]

euri[/url]":1m2sl4eh]You can get a 3.5 license through the 60-day evals available on VMware Partner Central if you have access

PM me if you don't

Sweet, now to go to that mess of a site and find it!

Will message you if I have no luck

[edit]

Found it!

Thanks euri, much appreciated. Means I dont need to faff around doing stupid stuff to get this environment over.

 

[ronelson]

The day I need Enterprise Vault to be working, it isnt.

Once more, with feeling! Email is not a document repository system!

 

[Danger Mouse]

PaveHawk-[/url]":1m1pu9eb]The day I need Enterprise Vault to be working, it isnt.

I need VMware 3.5 Enterprise license files that exist in the damn store. I have to migrate a client from 3.5 to 5.x with new physical hosts. I wanted to setup and use svMotion or even vMotion so I could get the job done quicker.

All the VMware .LIC files are in my EV archive.

The physical host that has the closed data partitions had a RAID controller + disk failure. I cant get to my damn EV archive

Along those same lines, it turns out my Data Domain wasn't actually backing up anything new for the last, say year or so

Probably an old DDOS bug combined with an old (now patched) Commvault Simpana v8 bug.

Had to shoot for a straight tape backup on the day prior to the data center move (actually did it at 3AM on Friday morning, move was scheduled to start at 7AM).

The move more or less went off without a hitch, although a few punchlist items remain, it's working.

And it's MY BABY! I made the point to my manager that he doesn't have anyone else to really take care of it right now, AND since most of my work is on the systems in there, AND noone else can really deal with those systems in there, I'm more or less taking ownership of the place and getting things set up right.

He had a troubled look on his face, but I think that's because he felt someone else could pick up the work besides me (not likely). Although there is one guy who could do it, he's got his hands full with a second career as autism specialist (his son has full autism spectrum disorder, nonverbal, etc).

Lessons learned:

-those extra depth/height racks? Totally necessary, especially if you've got any inclination to do cable management as well as easily mount the PDUs, etc

-inevitably, you'll think you have more than enough of the short power cables, only to run out of the long ones you need instead

-get the freaking dust cleaned out first, to prevent issues with wonky fiber connections as well as construction dust in the eye causing injury

-It's almost inevitable something will get forgotten in the end, hope that it's something relatively painless (ILO/IPMI switch wasn't patched)

-Vendors use too many custom screws for racking their gear. Design for and use the fricking standard stuff from the rack vendors, plz

 

[PaveHawk-]

ronelson[/url]":2n77fbpa]

The day I need Enterprise Vault to be working, it isnt.

Once more, with feeling! Email is not a document repository system!

The in-house file server that stored all the data I needed fell over about 3-4 months ago, because it was running RAID5 and wasnt monitored (to detect errors in the first place).

All I have left, now, is my emails and EV. Alas, the same issue occurred and disks failed in the server hosting the EV archives.

I'm dealing with incompetent people here, when I was running the internal infrastructure I knew about failures before they had any impact and mitigated them. Since I had it taken off me, our environment has gone to shit.

I also found out that I have a rat in my team. We dont deploy cable management arms, based on prior history. This guy believes we're all wrong, I put it to vote and excluded my own vote and the results were 3-1. Still, he went around me and went to another team to get them to exert pressure on deploying cable management arms.

The whole issue stems from the fact that the Support team (well, the senior in that team) is too lazy for his own good. He wants to perform work on a server with minimal effort possible to himself. You can put a host in maintenance mode, power it down and perform your works unplugging 4 TwinAx/Fibre cables, ethernet for DRAC and the 2 power cables. Apparently thats too hard. We must use cable management arms.

I really dont like people who have their own agenda and put it above the team. Especially when done in an attempted covert manner.

 

[sryan2k1]

unplugging 4 TwinAx/Fibre cables, ethernet for DRAC and the 2 power cables.

Sounds like our R810's. The whole thing takes like, 10 seconds.

 

[Rick25]

Management arms are evil. We banned ours years ago. Velcro and short power cables work just fine.

 

[Danger Mouse]

Rick25[/url]":38ghk1i0]Management arms are evil. We banned ours years ago. Velcro and short power cables work just fine.

Yup, no management arms in my new data center.

I did preserve one horizontal cable management piece to hang up all the various cables from the durn Celerra, but that was it.

 

[kperrier]

Apparently there is a project management methodology in which the PM just has a list of tasks, waves their hand and says that none of the tasks can be done in parallel and not assign a duration for any of them (there are well over 100 of them) and say that it will take 12 hours for the go-live to be complete.

 

[Technarch]

That's the one my boss learned, apparently.

 

[PaveHawk-]

Comedy. I didnt think this through. The client wants vSphere 5.1 and not 5.0u2. ESX 3.5 isnt supported on VC 5.1

 

[Barmaglot]

That's the part where you do a staged upgrade.

 

[PaveHawk-]

Barmaglot[/url]":1s7t3q5h]That's the part where you do a staged upgrade.

I found an email where I was right. They originally wanted 5.0u2, but changed their mind to 5.1 and forgot to tell me. In return, they've agreed for a full shutdown of their environment to allow me to migrate the VMs (its only ~10 VMs). The new host is hooked up to the old iSCSI SAN and I've got the new hosts in trial mode. I'll unregister VMs, register them on the new host and simply svmotion them.

Oh yeah, this is all new infrastructure. From 1st release R610s/MD3000i to 12g R620s/MD3220 SAS.

Quick calc shows that I can run their entire infrastructure on 0.35 of one of the new hosts.

Oh and Dell screwed up, I specced out for E5-2640s and got E5-2650s. First time its EVER worked out in my favour.

 

[Klockwerk]

TIL that vmware hosts really, really like holding onto mac addresses, to the point that if you move a network card from one host to another host in the same management VLAN, you're going to have a bad time *and* freak out the network guys as well.

They even have a KB article on it. It's 'by design', 'sucks to be you', and 'oh, if you don't want that to happen there's some stuff you can do. You'll have to reboot though afterwards'.

Then it got interesting. On top of their own range, they're using Xerox's first ever range when autogenerating mac addresses.

 

[sryan2k1]

Link to the article? I'm wondering if that can explain some issues I've seen in the past with our 10Gb NICs.

 

[Dqua]

sryan2k1[/url]":1yh8wam8]Link to the article? I'm wondering if that can explain some issues I've seen in the past with our 10Gb NICs.

Shoot me a PM, I uncovered a 10Gb NIC problem inadvertently months ago that is still on going. Not MAC/ARP related though.

 

[smwht]

TIL labels are important on disk shelves.

Also, controller cache can be a lifesaver when you pull the wrong drive

 

[hawkbox]

TIL that you actually can pay list price for IBM hardware. The people who were running this department before were completely pants on head.

 

[dredphul]

hawkbox[/url]":2a91eb88]TIL that you actually can pay list price for IBM hardware. The people who were running this department before were completely pants on head.

You can pay above list price if you want a huge Oracle server replaced yesterday (after ignoring multiple requests and warnings from Operations that the current server is too small for peak traffic in production).

 

[hawkbox]

dredphul[/url]":gpxt1wij]

hawkbox[/url]":gpxt1wij]TIL that you actually can pay list price for IBM hardware. The people who were running this department before were completely pants on head.

You can pay above list price if you want a huge Oracle server replaced yesterday (after ignoring multiple requests and warnings from Operations that the current server is too small for peak traffic in production).

Well this was willingly and without negotiation. They literally were shown the list price and said "Ok".

 

[M. Jones]

dlp[/url]":3bfjl5io]

afidel[/url]":3bfjl5io]

dlp[/url]":3bfjl5io]
Whenever there's any activity, usually the task manager will show system idle at 95% but CPU usage at 12% or something odd. I'm used to those numbers lining up usually.

That sounds like a precision/clock skew issue where you're getting different numbers reported at slightly different times shown on the screen at the same time. Do you have the VMWare tools installed? Also what is the CPU %RDY for the VM you are monitoring?

Yes to VMware tools.
%RDY seems steady at 0.20 to 0.21.

You may be right about it being a timing issue. When mostly idle, it seems accurate. I've just never noticed that before.

On the subject of clock skew: VMware currently recommend that VMware Tools not be used to synchronize time between host and guests. This represents a change from only a couple of years ago, at which time VMware recommended clock synchronization through VMware Tools.

• http://kb.vmware.com/kb/1318 # ‎Timekeeping best practices for Windows
• http://kb.vmware.com/kb/1006427‎ # Timekeeping best practices for Linux, with preferred kernel settings

 

[Technarch]

hawkbox[/url]":3lj0jgmr]
Well this was willingly and without negotiation. They literally were shown the list price and said "Ok".

My sales rep friend would very much like the contact info for your department.

 

[hawkbox]

M. Jones[/url]":3rr2geal]

dlp[/url]":3rr2geal]

afidel[/url]":3rr2geal]

dlp[/url]":3rr2geal]
Whenever there's any activity, usually the task manager will show system idle at 95% but CPU usage at 12% or something odd. I'm used to those numbers lining up usually.

That sounds like a precision/clock skew issue where you're getting different numbers reported at slightly different times shown on the screen at the same time. Do you have the VMWare tools installed? Also what is the CPU %RDY for the VM you are monitoring?

Yes to VMware tools.
%RDY seems steady at 0.20 to 0.21.

You may be right about it being a timing issue. When mostly idle, it seems accurate. I've just never noticed that before.

On the subject of clock skew: VMware currently recommend that VMware Tools not be used to synchronize time between host and guests. This represents a change from only a couple of years ago, at which time VMware recommended clock synchronization through VMware Tools.

• http://kb.vmware.com/kb/1318 # ‎Timekeeping best practices for Windows
• http://kb.vmware.com/kb/1006427‎ # Timekeeping best practices for Linux, with preferred kernel settings

Yeah that blind sided me at my new job.

 

[Klockwerk]

sryan2k1[/url]":1nxtiwhl]Link to the article? I'm wondering if that can explain some issues I've seen in the past with our 10Gb NICs.

vmk0 management network MAC address is not updated when NIC card is replaced or vmkernel has duplicate MAC address
http://kb.vmware.com/selfservice/mi...nguage=en_US&cmd=displayKC&externalId=1031111