That is incorrect, at least for iSCSI. Not sure about NFS. It just isn't exposed in the VI client.
what did you learn today?
- Thread starter chalex
- Start date
- Status
That is incorrect, at least for iSCSI. Not sure about NFS. It just isn't exposed in the VI client.
I'm pretty sure I did storage vmotion from FC to NFS on 3.5 when we switched our storage vendor
Storage VMotion works on NFS as well. I only had 3 until recently:
http://lewaness.wordpress.com/...tion-gui-in-vcenter/
Works awesome.
I spent an hour today fighting with the "export-mailbox" cmdlet on Exchange.
It seems there is a bug somewhere that makes it so you can't export to a PST file if you use the -startdate and -enddate options. However you can use those option to filter the contents into another mailbox and then dump the entire contents of that mailbox out to a PST file. --
--
It seems there is a bug somewhere that makes it so you can't export to a PST file if you use the -startdate and -enddate options. However you can use those option to filter the contents into another mailbox and then dump the entire contents of that mailbox out to a PST file. --
--I learned that if you have a polycom system and you are trying to troubleshoot why international ISDN video calls can not come in or out, it will take 2 weeks of back and forth finger pointing between your PBX vendor and your PRI provider before they finally figure out that A: maximum digits needs to be changed and B: the PRI vendor has a PIC code sending all international D-channel traffic over to MCI because they have no terminating agreements, and that they only applied that PIC code to 1 of your 2 PRI/trunk groups, so that when it hit the first PRI/trunk group and found it busy, and rolled to the next, there was no PIC code to route it to MCI.
and also C: certain polycomm and picturetel units will not communicate with each other. even though i'm pretty sure polycomm ended up aquiring picturetel.
That was a joyous task.
I also learned that even though the documentation says you can use ESX based VMware guests to create a riverbed RSP win2k3 package so you can have a virtualized win2k3 server on your branch office steelheads (gonna be print servers, EPO and WSUS repositories, i hope), that they are actually full of sh!t and only VMware server 2.0 guests will work to make the RSP package.
Dammit, MONOLITHIC disk -- --
i also am continually reinforced taht riverbed has one of the best support orgs out there, and that i can forgive one piece of bad info in one of their docs for the other 30,000 times they have pulled my ass out of a firey inferno.
it was a day of culminations today.
i also learned that if you go to boston during the thanksgiving weekend and drink "beer punch" at cambridge commons with your cousin and his lackeys, you will come back with the cruddiest sinus infection ever. swine flu got nothing on this crap. and no sick day in sight.
and also C: certain polycomm and picturetel units will not communicate with each other. even though i'm pretty sure polycomm ended up aquiring picturetel.
That was a joyous task.
I also learned that even though the documentation says you can use ESX based VMware guests to create a riverbed RSP win2k3 package so you can have a virtualized win2k3 server on your branch office steelheads (gonna be print servers, EPO and WSUS repositories, i hope), that they are actually full of sh!t and only VMware server 2.0 guests will work to make the RSP package.
Dammit, MONOLITHIC disk --
--i also am continually reinforced taht riverbed has one of the best support orgs out there, and that i can forgive one piece of bad info in one of their docs for the other 30,000 times they have pulled my ass out of a firey inferno.
it was a day of culminations today.
i also learned that if you go to boston during the thanksgiving weekend and drink "beer punch" at cambridge commons with your cousin and his lackeys, you will come back with the cruddiest sinus infection ever. swine flu got nothing on this crap. and no sick day in sight.
I love that place and that stuff. Circus Boy poured into lemonade and vodka? Yes please. If you want to drop a serious couple bucks though, next time you're up here, is Lord Hobo on Hampshire St. Frickin fantastic selection of draft belgians, like you've never seen.
Although I walk by it every day to and from work, I have yet to go to Lord Hobo (or Pickwick Ale, as I call it). I was never a fan of the B-Side, and I think some of that lingering feeling has stuck.
It's a new owner, the guy who owns Moan and Dove in Amherst and the Dirty Truth in Northampton. They still have good cocktails, but the emphasis is on the crazy-good beers. It's expensive, no joke, but eminently worth it.
Thank you for this --
-- I only checked this thread and saw your post after migrating 4 VMs out of 12, but this is definitely helpful. I installed the plugin and can confirm that it works on 3.5 with iSCSI and NFS. Hopefully I won't run into problems with free space for the snapshots it creates - those tiny individual datastores only have a gig or two free - but with system volumes, at night, it shouldn't be an issue.
I've learned once again that Kaspersky is the business. New virus, Kaspersky has the definition. BitDefender's online scanner heuristically found it. Still waiting on McAfee to come up with something. 2 hours later. Go Platinum support!
-=Whittey=-
-=Whittey=-
Sounds oddly familiar, had a similar problem in 2001 with SBC, they somehow dropped LD service from the second B channel on line 3 of a 6 line bundle in some kind of billing snafu. We could connect at 256Kbps but the quality was awful, 384+ connections dropped on negotiations. Took 3 truck rolls, an ISDN protocol analyzer and finally a tech who decided to call a remote loopback phone number on each B channel before they figured it out.
That our IT security folks can create duplicate DNS entries using the IP address we use for our production database, take down our application during the middle of the batch window (causing us to miss the SLO by 1.5 hours), and suffer absolutely no consequences.
Not to beat a dead horse here but I'm rather sure vMotion works with iSCSI and 3.5. If it does not, someone needs to tell my little 3.5 cluster it's doing something it cannot. I've moved live disks lots of times over iSCSI. There is just no VMware provided GUI to do it. A third party plug-in adds a simple GUI to the situation.
Same. PacBell and ATT did this to me at different times. ISDN most be fully misunderstood everywhere now. --
--Originally posted by erratick:
Same. PacBell and ATT did this to me at different times. ISDN most be fully misunderstood everywhere now. --
At my current employer we moved our partners to H.323 over IP because once we setup our gateway and got a local provider to setup their end it has *just worked* and as a bonus it's significantly cheaper to operate than a whole bunch of international LD calls on ISDN lines =)
I once spent months troubleshooting dropped video teleconference calls... international long distance ISDN from Maryland to the UK. Dealing with one telco is hell enough; in this case our calls went from Verizon to Sprint, across the Atlantic to BT and then to Cable and Wireless. Naturally, everyone pointed their fingers at their neighbor. Eventually we got Sprint to find an issue with one of their transatlantic trunks, and everything worked reliably after they fiddled with the routing.
Just curious, since we have platinum support as well. Did you guys get the email about reduced hours for Thanksgiving last week? That is total bullshit - that is when we need the support the most, when half of our workforce that has experience with the product is unavailable. Of course, our company buys the lowest level Platinum contracts (or last year's platinum level and pays extra for the sub-par level) so I am curious if it is just us or their standard support level.
Yah, regular (daytime phone call) platinum support was closed. You could still call your SAM and have them get someone intelligent on the line if there was an issue. It's better than HP's "we're taking the last couple weeks off" policy.
So now we're at 15 hours mark. Two SAM's, voicemail for the overall account manager and very little. I have an extra.dat. I have a name. And I have an email that says "Attached is the extra.dat for this detection. Please use the beta.dats for registry cleanup." No more information. Solid platinum...
-=Whittey=-
Ha, okay, same thing. That is not platinum in *my* book, but since I do not matter... FWIW, our HP deals are 24/7/365 at 4 hour and we have never had an issue with it. McAfee are the only people that take off for the holidays at the top tier support level.
I've learned that attempting to install Exchange 2010 on a computer whose clock is more than 10 minutes off of the DCs will result in a hung install process.
I learned that it's possible to create a RAID10 array on an R710 consisting of two 73GB 15K disks and two 300GB 10K disks. I learned this because we received it from Dell that way.
I learned that working at a place where management lets the developers dictate how the production environment should look like instead of leaving that to the ITO department will inevitably lead to endless pains for the latter. Not to mention for the company as a whole.
I learned that having nationwide on line elections and choosing the biggest provider in the country to do the networking will not guarantee that some idiot doesn't forget a traffic shaper on the cluster firewall.
This resulted in practically having 16MBits down and 1MBit up, with 3500 clients pounding the cluster demanding for a reply.....
This resulted in practically having 16MBits down and 1MBit up, with 3500 clients pounding the cluster demanding for a reply.....
I hear ya. "No, we cannot operate our single-threaded application infrastructure without 10 x 4 vCPU VMs running on 3 dual-quad-core blades."
Hello CPU ready time.
Sometimes their LUNS and partition makes no sense. Like they throw darts to decide how to carve it up. Oh well, rather stage the hardware from scratch myself anyway.
I learned today that if you delete a user's mailbox in Exchange 2007, it will also delete the associated AD account.
Trying to determine what line cards, supervisors and chassis work together for Cisco gear can be a nightmare; especially when two pages on Cisco's site contradict each other.
http://www.cisco.com/en/US/pro...900aecd8035cf2b.html :: "The supervisor engine IV is compatible with the classic Cisco Catalyst 4500 and the E-series Cisco Catalyst 4500 Line Cards."
http://www.cisco.com/en/US/pro...ucts_Data_Sheet.html :: "Cisco Catalyst 4500 E-Series Line Cards -- Supervisor Engine 6-E, Supervisor Engine 6L-E,"
Which is it Cisco?
http://www.cisco.com/en/US/pro...900aecd8035cf2b.html :: "The supervisor engine IV is compatible with the classic Cisco Catalyst 4500 and the E-series Cisco Catalyst 4500 Line Cards."
http://www.cisco.com/en/US/pro...ucts_Data_Sheet.html :: "Cisco Catalyst 4500 E-Series Line Cards -- Supervisor Engine 6-E, Supervisor Engine 6L-E,"
Which is it Cisco?
Ouch, you got bit by the "Remove" command. --
-- Have no fear, everyone does. It doesn't help that in Exchange 2000/2003 "Remove" meant "delete the mailbox, leave the user alone" and then they changed it in 2007/2010.I'm surprised that's all that broke. Kerberos is *not* happy if the time delta is >5 minutes. --
--Nope. Didn't know it existed (and we'd get our VAR to double-check for us); playing with it now, and while the SupIV is compatible with the chassis, it's not compatible with a sample of IOS that I've picked. Urghh.
Edit: Enable Configuration Guidance. Red bars on things that don't play together.
That you don't fsck with database server unless the Netflow Analyser machine that uses it is shut down. Go-go gadget database re-init!
That it's near impossible to debug a HP network team on a full flight production server. I have between 8pm and 9pm each day to figure out why the bitch keeps dropping connections, and I'm lucky to get that. At least the HP support site has a "Submit a ticket", or I'd have to wade through the support phone line.
"Wassat honey? Oh, you bought Christmas gifts? Nice! Who's this fo - er wha? Not enough money left for booze this weekend? Nice."
That it's near impossible to debug a HP network team on a full flight production server. I have between 8pm and 9pm each day to figure out why the bitch keeps dropping connections, and I'm lucky to get that. At least the HP support site has a "Submit a ticket", or I'd have to wade through the support phone line.
"Wassat honey? Oh, you bought Christmas gifts? Nice! Who's this fo - er wha? Not enough money left for booze this weekend? Nice."
SMARTS In-Charge is a total piece of shit. I knew it was a piece of shit, but I just learned that it is 100% shitty.
I import a seedfile that has the new SNMP string for 100+ devices. They all fail, yet snmpwalk works fine. I experiment by deleting a device and then importing it - works fine. So, I yank the devices out and reimport them from scratch. SMARTS locks up importing a single device for two hours. Suddenly it dumps a ticket for every modified device with a down trap. HOW? I just deleted them, there is no up state established! DIE DIE DIE!
I import a seedfile that has the new SNMP string for 100+ devices. They all fail, yet snmpwalk works fine. I experiment by deleting a device and then importing it - works fine. So, I yank the devices out and reimport them from scratch. SMARTS locks up importing a single device for two hours. Suddenly it dumps a ticket for every modified device with a down trap. HOW? I just deleted them, there is no up state established! DIE DIE DIE!
Eve ate my rant about the Exchange 2007 installer running perfmon invisibly in the background. -- --
Long story short.. somehow a brand new Server 2008 ENT build would crash out perfmon everytime you added counters to it.
It took about 3hrs with MS Support to figure out that is what was crashing the Exchange installer.
Then another hour trying hotfixes, regedits and resetting the perfmon counters to get the Exchange 2007 installer working again.
About 15mins later I had my 5th node back and running in my cluster.
Good news:
While I had this downtime I ran some offline defrags and cleaned up ~900Gb from my Exchange stores -- --
[edit]
Here's the hotfix we ran: http://support.microsoft.com/kb/961382 My Exchange nodes have about ~400 multipathed LUN's presented to them, which seems to be what caused this.
--Long story short.. somehow a brand new Server 2008 ENT build would crash out perfmon everytime you added counters to it.
It took about 3hrs with MS Support to figure out that is what was crashing the Exchange installer.
Then another hour trying hotfixes, regedits and resetting the perfmon counters to get the Exchange 2007 installer working again.
About 15mins later I had my 5th node back and running in my cluster.
Good news:
While I had this downtime I ran some offline defrags and cleaned up ~900Gb from my Exchange stores --
--[edit]
Here's the hotfix we ran: http://support.microsoft.com/kb/961382 My Exchange nodes have about ~400 multipathed LUN's presented to them, which seems to be what caused this.
Figures... it lets me post that^^ just fine. Grrrr.
Been awhile since I posted here, have some catching up to do!
Please note, if you have a vendor running their 'Emergency Department Information Systems' from an Oracle database on a Win2003 cluster, they probably don't know what they're doing.
They will somehow pick the one time that a Oracle backup is running (that no one knew about/scheduled??) to fail the database over.
This will:
Corrupt the database
Cause a 10min outage to run 2.5Hrs
Require a complete restore of the 20Gb .ora from RMAN
Cause 5 hospital ER's to fall back to paper on a Friday night
And end a ~$15 million/50 hospital rollout at 5 hospitals and a lawsuit.
They will also under-spec their required enviroment by almost 70% and cause a ~$200,000 budget overrun in server hardware alone.
Fun times!
Been awhile since I posted here, have some catching up to do!
Please note, if you have a vendor running their 'Emergency Department Information Systems' from an Oracle database on a Win2003 cluster, they probably don't know what they're doing.
They will somehow pick the one time that a Oracle backup is running (that no one knew about/scheduled??) to fail the database over.
This will:
Corrupt the database
Cause a 10min outage to run 2.5Hrs
Require a complete restore of the 20Gb .ora from RMAN
Cause 5 hospital ER's to fall back to paper on a Friday night
And end a ~$15 million/50 hospital rollout at 5 hospitals and a lawsuit.
They will also under-spec their required enviroment by almost 70% and cause a ~$200,000 budget overrun in server hardware alone.
Fun times!
WhiteHat - are you running E-series cards, and if so, do you get the full 24 Gbit/s switching for their slot? Cisco's docs read as 'It works/(might work), but they run at 6 Gbit/s.'.
- Status