what did you learn today?

[PaveHawk-]

This isnt so much something I learnt, more about something I hate: Tape pass-through in VMware.

2+ years ago, our Business Development team decided that a single VMware host + backup server was too costly, and that we needed to create a single server solution using a virtual backup server. The end result was tape pass-through using the only SCSI controller on VMware's HCL at the time and an LTO3, then presenting that to a VM and letting Backup Exec have at it. Unfortunately, I got tasked with upgrading their formerly ESX 3.5u3 server to vSphere 4.1u1 server.

Its a PowerEdge 2900 with an E54xx CPU, so no VMDirectPath for me I WANT A DEDICATED BACKUP SERVER GODDAMNIT. Still trying to get this thing working after 4 days.

I learnt that sometimes I hate my BusDev team.

 

[PaveHawk-]

Go E2010, you'll be thankful for it.

 

[PaveHawk-]

That the previous IT vendor for a client of mine are morons. They deployed an IBM DS 4700 Express using 4Gb FC on a pair of Brocade switches. No zoning whatsoever.

I'm not all that up to speed on FC, though it appears I'm about to get a crash course, but I figure you'd want to use zoning to limit access and prevent issues of loss of data. Currently the backup server hits the IBM tape library over FC, the same fabric, but as there is no zoning the ESX hosts can see the tape drive as well.

That might not be an issue, but the fact that I can see the volumes in diskpart/disk manager on the backup server poses a few issues for me. I'm waiting for one of the other engineers who arent aware to go in there and go "hey wow, 1TB of space, I could use that for B2D!!" then mount and format the volume.

That'd make my day.

 

[PaveHawk-]

One good thing came out of the vSphere licensing changes. We now have Storage vMotion and DRS. Will be so nice to be able to move VMs among datastores. :bigdumbgrin:

Well, in news related to this, managed to convince the MD that we shouldnt go blade (we couldnt even fully populate a single chassis) and that rack would be better for us. As a result, I got approval to buy and have received some of this:
- 4 x Dell R710 (in time for me to whack VS5 on it to have a gander)
- 2 x Cisco 3750X's (to go with our 3750G stack)
- EMC VNX 5300 (EMC is literally 1 floor up from us, so I got one of the presales guys to come out and help me with setup etc)
- VS5 Ent+ licensing (As a partner, we get 4 Ent+ licenses free, so we bought the other 4 and all the maintenance)
- A HP TippingPoint IDS on loan, with us likely buying it

What I didnt get:
- 10GbE (Switches and cards, my life would have been easier)
- New C2960S access stack (this was hoping for too much)
- Palo Alto PA2020, because its cooler than anything Cisco in the lower end of the market
- 10KVA UPS (each floor only gets 100 amps, they've built the building with floors grouped together, so we're on an 800amp circuit with 8 other floors and we share our 100 amps with another tenant. To run another feed up from the basement is around $30-50k, potentially more since they think elements of the electrical environment may not come back up if its powered down. All I wanted was a single 63 amp feed.)
- More disk in the EMC VNX 5300, specifically I wanted EFD as my top tier and not SAS - but the cost is still too high. Maybe I might get 1TB later this year in EFD.
- More RAM in the servers (they're all 96GB)
- A proper rack and PDUs
- Aircon guys to come out and fix their fuckup (its a server room, not a DC, but why is the cold air vent pushing air down the back of the rack and why is the hot air extraction vent in the front of the rack?)
- A sledge hammer (for use on staff who decide to "change" something in the environment because they thought it was the right setting versus coming and checking with me before touching anything - so help me god, between RBAC, ADSIEDIT and vCenter permissions, I am locking shit down. No more fucking around people.)

 

[PaveHawk-]

Restricted space behind the racks. Though I just managed to get the MD to go "fine, get a proper quote for it" to get that fixed. Said "well, we put all this new gear in and if it dies due to overheating its not my fault ok?".

 

[PaveHawk-]

I learnt today that I've learnt a lot from scorp, directly and indirectly.

I'm sitting at AU TechEd at an Exchange 2010 virtualization session and frankly I'm not learning anything new. Bored as hell here...

 

[PaveHawk-]

Mebbe its me giving the talk? Heh... I wish I was in AU right now. Sorry that you're bored though, that sucks.

Well seeing how I know your name (given you asked me to relay something to Scott Schnoll and all), I'd be able to pick you out

You should see if you can volunteer for talking at TechEd AU, would be a good introductory event to talk infront of around 60-80 people (if Andrew Ehrensing's sessions are any indication, its closer to 100). Since I'm not getting pushed to a role involving more meet-n-greet stuff, I'll be here regularly from now onwards even though its actually not really that useful for me.

That said, they do have a 400 session about AD, virtualisation and cloning that could be good.

TechEd sucks when MS dont have anything new to really push hard (they're sort of pushing Denali, sort of pushing all the System Center stuff - but its a bit half hearted). Last year they had Lync 2010, E2010 SP1 and a few other things that made it worthwhile from an infrastructure POV. At least the devs get to see all the WinPhone stuff, as well as the big BI push that MS AU have been doing. Oh and SP2010 (though, that only seems big since our new hire is speaking here, since he's a VTSP).

 

[PaveHawk-]

(if Andrew Ehrensing's sessions are any indication, its closer to 100).

No way! Andrew is an awesome dude, he usually has high level stuff. Maybe he was told what to focus on and couldn't do his usual stuff.

Should clarify, I mean he's getting 100 people in his sessions when the room is really only designed to fit around 60-80. When its a slightly lighter material set, he makes it more fun than the dry boring stuff. Even with the dry technical stuff, he still manages to get through a lot without making it seem like you want to go to sleep.

Taht said, I brought supplies; a can of redbull and a bottle of water for the post-lunch crash.

 

[PaveHawk-]

Oh ok *phew*... I misunderstood thinking 100 level content.

Speaking of content, here I am sitting in a session waiting for it to start. The speaker today is...Andrew Ehrensing. He says hi btw

[edit]

Its EXL304 - Load Balancing with Microsoft Exchange Server 2010

I'm here to kill some time though, since I'm waiting for the SCVMM 2012 stuff for the most part.

 

[PaveHawk-]

In TechEd.AU, Colin Lee just said that PF access in E2010 is still direct to the mailbox role and not through CAS. Hm, didnt know that one!

 

[PaveHawk-]

People still burn in servers? Don't they do that at the factory?

Nope, if I get a HP server it arrives in about 4 trillion boxes (they need a box for the box for the box for the box for the box for the box etc), so I have to remove everything from the boxes, curse HP and their logistics systems, pray I got all the right parts, pray I got ALL the parts, assemble it and hope to god that it works properly without errors. No way do I put that out without at least a memtest run...

 

[PaveHawk-]

Get thee to a decent VAR, CDW does all that for us for a fairly nominal fee (way less than our volume discount on most servers).

Last I checked, CDW isnt in AU - and the fees I get charged for that arent nominal either (usually 5-7% cost of server) with the current vendors.

Hopefully we're getting a junior who'll be tasked to do all the build work on servers so I can spend my time actually doing the higher end stuff that I'm tasked to do...

 

[PaveHawk-]

People still burn in servers? Don't they do that at the factory?

Nope, if I get a HP server it arrives in about 4 trillion boxes (they need a box for the box for the box for the box for the box for the box etc), so I have to remove everything from the boxes, curse HP and their logistics systems, pray I got all the right parts, pray I got ALL the parts, assemble it and hope to god that it works properly without errors. No way do I put that out without at least a memtest run...

If this is normal practice for HP, I can't see why they'd be so popular.

Next DL3 or DL5 that I assemble, I'll take a photo of the crate of boxes or something. In fact, Dell did a comparison of their M1000e + blades vs C7000 + blades and whatever IBM's equivalent bladecenter is. In it, part of the test was to assemble, they took photos of it - I'll have to dig it up.

Not saying that HP is better than Dell or vice versa, but when its me doing the build I start appreciating Dell's single box philosophy.

 

[PaveHawk-]

This is not standard practice. This is your VAR doing this to you. I have been buying HP servers for 10 years before I worked for them and the truth is you should be getting it all factory integrated. If you need it overnight and shipped directory from your var maybe I could see this but I have been to the factory where this gets put together. It take 20-30 minutes from when the order hits the factory floor to having a server sent over to burn in. The dell picture is very misleading... I would say 99% of blades and server ship in a single crate ready to plugin. You can order it piecemail by why in you would I have no idea. I had one customer bypass their account rep and went to a VAR who tried this and I guarantee you 14 pallets later it never happened again.

We've got 2 suppliers who provide us HP gear, both have the same requirement for me to pay to get it assembled - no HP option to get it in a single box. That said, its not THAT much of an issue, the gear is still good so its taking a relatively small point and blowing it out of proportion.

I'll put the issue back on to my logistics coordinator and ask them to push our reps in clarifying the build situation; but I wonder if its an AU market thing as well.

 

[PaveHawk-]

Thumbs up to Netapp for missing their SLA by 4 hours?

Better than EMC's 7 hour call back for a Sev1 issue on a NS-960 last Thursday. That was after 4 emails to our support rep asking for updates.
That's why we pay them millions of dollars a year in support for. The great support.

You always have so much luck with EMC...

 

[PaveHawk-]

That sometimes my clients are fucking idiots.

Client is deploying a new point of sale product that uses RHEL/JBoss tying into MSSQL. They want to learn the front end/middleware stuff so they asked me to spin up a RHEL5.6 (its only compatible up to 5.6 at the moment) machine, which I did.

They're also pro linux and anti MS (at least for licensing costs). Their onsite tech just emailed me asking for a GUI on the new RHEL server.

Right.

 

[PaveHawk-]

DOH. I've had customers ask for that. They purchase a centos or RHEL VPS or dedicated server and the first question is: How do I get to the desktop?

Even funnier, they want it for Nagios and I just checked bash_history and had a nice chuckle at the attempt to install Apache. Downloaded the .tar.gz from an Apache mirror site, then tried to run yum on it.

 

[PaveHawk-]

If you haven't already heard of it? A Periodic Table of Visualization Methods

That is both awesome and funny at the same time.

 

[PaveHawk-]

Even funnier, they want it for Nagios and I just checked bash_history and had a nice chuckle at the attempt to install Apache. Downloaded the .tar.gz from an Apache mirror site, then tried to run yum on it.

On a related note, developers don't really understand why you're not thrilled to install and support Tomcat7, but when presented with the right incentives are perfectly willing to listen to reason.

From .bash_history:

mv Httpd-2.2.20.tar.gz /user/local/: Cd /user/Local/tar
mv Httpd-2.2.20.tar.gz /user/local/: Cd /user/Local/tar -xzvf httpd-2.2.20.tar.gz
mv Httpd-2.2.20.tar.gz /user/local/: Cd /user/Local/tar --xzvf httpd-2.2.20.tar.gz
mv Httpd-2.2.20.tar.gz /user/local/: Cd /user/Local/tar xzvf httpd-2.2.20.tar.gz
mv Httpd-2.2.20.tar.gz /user/local/: Cd /user/Local/tar -xzvf httpd-2.2.20.tar.gz
mv httpd-2.2.20.tar.gz /user/local/: Cd /user/Local/tar -xzvf httpd-2.2.20.tar.gz
--
mv httpd-2.2.20.tar.gz /user/local/: cd /user/Local/tar -xzvf httpd-2.2.20.tar.gz
mv httpd-2.2.20.tar.gz /user/local/: cd /user/Local/tar  httpd-2.2.20.tar.gz
mv httpd-2.2.20.tar.gz /user/local/:
cd httpd-2.2.20
mv httpd-2.2.20.tar.gz /user/local/; cd /user/Local/tar -xzvf httpd-2.2.20.tar.gz
mv httpd-2.2.20.tar.gz /user/local/; cd /user/Local/tar -xzvf httpd-2.2.20.tar.gz cd httpd-2.2.20

mv httpd-2.2.20.tar.gz /user/local/; cd /user/nagios/ -xzvf httpd-2.2.20.tar.gz cd httpd-2.2.20

mv httpd-2.2.20.tar.gz /user/local/; cd /user/nagios/downloads/ -xzvf httpd-2.2.20.tar.gz cd httpd-2.2.20
mv httpd-2.2.20.tar.gz /user/nagios/; cd /user/nagios/ -xzvf httpd-2.2.20.tar.gz cd httpd-2.2.20
mv httpd-2.2.20.tar.gz /user/nagios/; cd /user/nagios/tar -xzvf httpd-2.2.20.tar.gz cd httpd-2.2.20
mv httpd-2.2.20.tar.gz /user/local/
mv httpd-2.2.20.tar.gz
cd ~/downloads
mv httpd-2.2.20.tar.gz /root/downloads
mv httpd-2.2.20.tar.gz /root/
cd /root/downloads
cd /root/downloads/tar -xzvf httpd-2.2.20.tar.gz
tar xzf httpd-2.2.20.tar.gz
tar xzf httpd-2.2.20.tar.gz
cd httpd-2.2.20.tar.gz
cd httpd-2.2.20.tar.gz
lynx http://httpd.apache.org/downloads.cgi
wget http://httpd.apache.org/downloads.cgi
yum install httpd-2.2.20.tar.gz
yum install httpd php
logon
nagios
n@gios
logon
exit
yum install httpd php
cd ~/nagios/downloads/httpd-2.2.20.tar.gz
cd ~/nagios/downloads/httpd-2.2.20
cd ~/downloads
cd ~/httpd-2.2.20
cd httpd-2.2.20
cd httpd-2.2.20.tar.gz
ogoff
cd ~/root
exit
mkdir /downloads
cd ~/downloads
wget http://httpd.apache.org
exit
sudo apt-get install apache2
sudo apt-get install apache2yast install apache2
yum install httpd php

su -1
/usr/sbin/groupadd nagcmd\
/usr/sbin/groupadd nagcmd
/usr/sbin/usermod -a -G nagcmd nagios
/usr/sbin/usermod -a -G nagcmd apache
/etc/init.d/httpd start
vi /etc/rc.local
rpm -e httpd mysql
mv httpd-2.0.55.tar.gz /usr/local/; cd /usr/local/
 tar -xzvf httpd-2.0.55.tar.gz
 cd httpd-2.0.55
 ./configure --enable-so
 make
 make install
mv httpd-2.0.55.tar.gz /usr/local/; cd /usr/local/
 tar -xzvf httpd-2.0.55.tar.gz
wget http://apache.mirror.aussiehq.net.au//httpd/httpd-2.2.20.tar.gz
mv httpd-2.0.55.tar.gz /usr/local/; cd /usr/local/
 tar -xzvf httpd-2.0.55.tar.gz
cd httpd-2.0.55
[root@support03 local]#
[root@support03 local]#
[root@support03 local]#
mv httpd-2.0.55.tar.gz /usr/local/; cd /usr/local/
 tar -xzvf httpd-2.0.55.tar.gzmv httpd-2.0.55.tar.gz /usr/local/; cd /usr/local/
 tar -xzvf httpd-2.0.55.tar.gz
om previous errors
[root@support03 local]#
om previous errors
[root@support03 local]#
mv httpd-2.0.55.tar.gz /usr/local/; cd /usr/local/
 tar -xzvf httpd-2.0.55.tar.gz
 cd httpd-2.0.55
 ./configure --enable-so
 make
 make install
mv Httpd-2.2.20.tar.gz
mv Httpd-2.2.20.tar.gz /user/local/
mkdir  /user/local/
mkdir ~/user/local/
mv Httpd-2.2.20.tar.gz /user/local/; cd /usr/local/tar -xzvf httpd-2.2.20.tar.gz
mv httpd-2.2.20.tar.gz /user/local/; cd /usr/local/tar -xzvf httpd-2.2.20.tar.gz
mv httpd-2.2.20.tar.gz /user/local/; cd /usr/local/tar -xzvf httpd-2.2.20.tar.gz
pgpk -a KEYS
gpg --import KEYS
 gpg --verify httpd-2.2.20.tar.gz.asc
4504934464c5ee51018dbafa6d99810d *httpd-2.2.20.tar.gz
 /usr/local/apache2/bin/apachectl start
yum install httpd-2.2.20
yum install httpd-2.2.20.tar.gz
exit
grep .^User. /etc/httpd/conf.d/httpd.conf
exit

Things worth knowing:
- RHEL 5.6 VM
- Username that is being used is "root"
- Its not registered with RHN as there is no license for it
- This is comedy gold

[edit]

I should add, after all that nothing has changed on the VM. I shouldnt laugh at this, but out of curiousity I googled "rhel 5.6 install apache" and got half a dozen links on how to do it.

 

[PaveHawk-]

For our internal infrastructure, we bought another pair of 3750's (3750X-48) to go with our existing pair (3750G-48). I'm not massively strong on Cisco gear by any means, but my co-worker and I managed to relocate some gear, install the switches and add them to the stack whilst both on minimal amounts of sleep.

Didnt help that I got home and in bed by 1230 in the morning, then came to work by 0715.

What I've learnt is that if I only have 4 hours of sleep on 2 consecutive nights, my eyes start burning pretty badly.

 

[PaveHawk-]

An upgrade of EMC Networker 7.6.1 to 7.6.2 has taken a week, 6 roll backs to 7.6.1 and now an engineering escalation for my SR. Sounds reasonable for a point release... right?

I actually keep a desktop a work that is solely for EMC to Webex in to fix Avamar and Networker. About once a month we 'find an issue' that requires 3-4 days of their support. Very complicated things like:
A minor update to their software
An explaination as to why Avamar is skipping entire drives but reporting 'Successful'
Troubleshooting why their NMM agent for Sharepoint suddenly stopped working
Getting an actual answer as to whether XenServer 5.5 is support on the NS platform
Why DataDomain can only restore @ exactly 20MB/s
etc..

Watching your experiences with EMC always has me worried with our engagement with EMC. Hopefully its not as horrific as it has been for you.

 

[PaveHawk-]

I feel like my heart is actually a bastard and trying to kill my liver, or at least give it a livorce.

Why is that term not on urbandictionary.com yet? Chop chop, get it done Soko. Needed it done yesterday!!

 

[PaveHawk-]

....that I have been ordered by senior management to let some farkin management consultant firm, send their rainbow colored signature block through our email system, even though the email client generating the email apparently has a corrupt or possibly virus infected document/email template.

They already tried to order me to work overnight on their problem, WHICH IS OUTSIDE OF MY ORG AND MY CHAIN OF COMMAND.

Yup, off I go to modify our farkin Exchange Server System, just so one consultant can send their multicolored signature block through. Who does that with email these days ANYWAYS?

IS THIS FARKIN GEOCITIES? AOL? GRAGGAGAGGHGHGHGHGHGHGHHLLLLEEE

Edit: The person's email client has a corrupt template file, causing the multi colored text to have weird/corrupt characters in it, triggering Forefront's content scan to flag it.

Ahh, so bad yet so funny.

 

[PaveHawk-]

EMC vs

Sanguy, you might want to put your responses BELOW the quoted material. Most of us read left to right, top to bottom, so it is slightly confusing. This ain't Outlook, ya know

Let's leave that particular flame war out of this discussion please. It's like Vi vs Emacs or Linux vs Windows or any other long list of pointless battles.

Netapp?

Figured I'd get a few more in there...

 

[PaveHawk-]

I learnt today that I dont understand EMC's licensing at all.

So we received our VNX 5300, we got the FAST and Local Protection Suite. Each enabler file is 140KB in size and comes on its own CD. Why? Why am I swapping CDs out to access these files so I can install it? Cant I simply just go to a website (NOT POWERLINK!!!!!!!!!!!!!!!!!) and download it? Surely that makes more sense?

 

[PaveHawk-]

Awesome. My brand spanking new VNX 5300 is making my day. Overnight, I come back and I'm poking around on the System tab of Unisphere and I find two issues:

1) Data Mover 2, Status: Out of service
2) Read cache is disabled on both of the controllers after FAST Cache was enabled.

Okay, I havent even done anything with the DMs ffs, its all plugged in but no configuration is setup (all disk is assigned for block; going to use the DMs later on when I can get more disk).

On the flip side, EMC's Beta Support site is miles better than PowerLink.

 

[PaveHawk-]

On the flip side, EMC's Beta Support site is miles better than PowerLink.

What is this thing you speak of... ?

https://supportbeta.emc.com/

I noticed it when reading Chad Sakac's blog, he mentioned the support beta site so I've been using that a little more. Search is still average, ended up PM'ing Pokrface for some tips on how to find stuff.

Read cache is disabled on both of the controllers after FAST Cache was enabled.

Yeah, I could have told you that. I had the Fast Cache enabler on our 5700 installed after everything else had been done, and it definitely left cache disabled.

I pointed that out to my local tech.; it is obviously a bit of a bug.

How'd they fix it? I havent yet logged a call regarding all of it yet since its not in prod, but I'm curious to know.

 

[PaveHawk-]

They didn't do anything, I just re-enabled the cache in Unisphere system properties, manage cache...

Well then, I have a problem - its greyed out.

Well, time to log another call.

 

[PaveHawk-]

Look at this:

Check memory allocation and I think you may find that memory is fully allocated; activating that feature does increase the footprint of the Flare.

So you can't enable read cache because there is no memory allocated to it. What you need to do is disable write cache, lower its allocation marginally, then allocate that space to read and re-enable both.

...or not...

You kill me. I worked that out about 45 seconds after posting, spent 5 mins to fix it and then spent the rest of the time berating myself at such a stupid oversight.

[edit]

Seriously, I'm fucking stupid for not actually thinking about that. Cant believe how fucking stupid I am over this one. ARGH. Is it wrong to have a few shots of alcohol (Herradura Tequila, good stuff btw) at 0915?

 

[PaveHawk-]

I learned that apparently VMware recommends disabling VAAI UNMAP now on all ESXi 5 hosts.

http://www.boche.net/blog/index.php...sues-recall-on-new-vsphere-5-0-unmap-feature/

There goes one really nice benefit for forking over all that extra cash. Don't know anyone bitten by any of the problems they identified, but it sounds like it's pretty universal and bad if they're going to patch it disabled globally. Geez.

Funny, I'm just migrating a bunch of machines onto 4 new vSphere hosts running VS5. That error occurred for us on 1 VM during migration, which was fun. No data loss, but it is timely...

 

[PaveHawk-]

I suppose that's one way of looking at it. Another way is that I work for a company/unit/division that does not invest much in me at all, so I invest in myself. When it comes to tools, training, and other forms of professional development, I'm pretty much on my own - so this is just one more expense I write off as a professional. I'm not married to this job or this company, but I also am not going to let the lack of support hinder my professional growth.

Besides, I like tech - so I have no problem with buying/building my own. It makes me a better consultant, which benefits me in the long run.

I bought a Macbook Air, the ethernet adapter, usb-to-serial adapter because the company I work for, the laptops we have for team use are pretty poor. I find its easier to buy my stuff and claim it on tax than to rely on the company, to a degree. It keeps them on their toes, so they're constantly talking to me to make sure I'm not going for a new job.

 

[PaveHawk-]

And we're finding out that seven CRAHs at 90-100% is much more efficient than eleven CRAHs at 50%. Thankfully we have capacity out the wazoo so we have a wide latitude for experimentation. And to be honest, it's fun trying to figure it all out. (God, I'm such a dweeb...

Well, its better to have and not need than to need and not have.

 

[PaveHawk-]

Ummm....this is the way we want it. It works well at all our sites and eliminates issues.

Disabling auto-negotiation almost always introduces issues.

I'm not a networking guy by trade, but I can negotiate my way around switches and routers. I have a C2821 that has a 4 port switch module, if I plug the switch module into my C3750 stack it throws up errors unless I hard set it to 100/Full on both ends. I quizzed TAC about it and I did get 2 different responses with 1 guy saying that I should always hard set to 100/Full between Cisco devices (I'm guessing that it doesnt apply to GbE gear) and another saying that I shouldnt.

And now I've completely forgotten my point, but I'll post this anyway...

 

[PaveHawk-]

That weird shit happens and I cant control it

My brand spanking new vSphere 5 cluster was humming along, until today when I noticed that HA and DRS are magically turned off. I say magically, because nothing in Tasks and Events indicates any changes other than when I turned DPM off entirely. I've got restricted access to the vCenter with only those with VCP able to do anything (the VCP restriction enabled me to limit the number to 3 people in total), but some service accounts require access until I can work out what they need (stupid vRanger).

I'm utterly lost; now I need to schedule a window so I can turn it back on.

 

[PaveHawk-]

Not sure why you need a window to turn HA/DRS back on.

This.

In reply:

Not sure why you need a window to turn HA/DRS back on.

Aside from general change control procedures, I would revert settings that had magically turned off during a window, too, in case the magic turns black.

This.

Things that magically change without any insight as to why, is something I'm not going to revert back during business hours.

Getting a window is easy for me, I just need to make sure I take into account the Western Australian guys who are 3 hours behind AEDST. So a 1730 change for me really is a 2030 change, since it has to be 1730 in WA.

(stupid vRanger).

The worst backup product I've ever dealt with, and we use Symantec products! It being that shitty combined with no budget, I've moved our vS 4 backups to using a few custom batch scripts and a VCB proxy. That works more reliably then vRanger ever did.

I love when it decides that disk 1 is no longer disk 1 and starts to backup some other disk that its randomly chosen. We're going to move across to Veeam, since thats more stable. Pity neither Veeam nor Quest think that spooling to tape is important and still require you to use some sort of tape backup application to backup your images.

[edit]

Oh and it is the storage profiles causing it, just tested it on a test cluster.

 

[PaveHawk-]

So, I passed my VCP5 exam today. Did the exam in 22 minutes including review. I did it without study or prep, completely blind and scored 311 out of 500. I'm not sure what its telling me, but I think it might be that the exam wasnt that hard.

Oh wait, I know, never review your questions. First answer is always right. I reviewed, changed, and am certain I lost score as a result (9 questions reviewed out of 85).

 

[PaveHawk-]

Congrats! Did you already have the VCP4?

Thanks.

Yep, did VCP4 a while ago (end of 09 I think).

 

[PaveHawk-]

Planning on doing the same thing. Studying is for pussies.

And congrats.

I'll pay that, since I passed! Thanks

In other news, I genuinely fucking hate Dell. Right at this very moment, I want to stab the Dell retard employees who came up with this whole SBUU + OM Repository way of updating the firmware of a server. You fucking retards, HP can do it with 1 disc so that means you should be able to.

Doesnt help that the drive is so finicky that it wont read all the Dual Layer discs that I put in it (nevermind that all the discs originate from the same batch and are all burnt via my machine at the same speed with the same application etc).

I seriously want to kill someone.

 

[PaveHawk-]

That's weird...I read a number of vSphere 4 books, had a year or two of experience, and still failed the VCP4 first try. Granted a lot of the questions I think I got wrong were large environment-based (small env. experience). Still haven't had a chance to re-take.

VCP4 tended to focus heavily on the min/max stuff versus the VCP5. I found the best way, for me at least, was to look at the question and answers and work out what was wrong. By working backwards, I deduced the right answers. Though, the more I think about it the more I realise I shouldnt have tried to speed through it, there are a few questions coming back to mind that I know I put the wrong answer for.

 

[PaveHawk-]

Citrix QA sucks.

http://forums.citrix.com/thread.jspa?th ... 0&tstart=0

You cant deploy XenDesktop 5.5 with vSphere 5 and Distributed Virtual Switches, cos it doesnt work.

Comedy gold.