And now I feel like I wasted an afternoon...sryan2k1":3ewfic0i said:
But thank you very much!
what did you learn today? (part 2)
- Thread starter continuum
- Start date
But thank you very much!
If all my napkin level calculations are correct, I got the Flops to run at least one of the GFS ensemble members on the vSphere cluster. If I could tie in all the faculty and student computers, I could probably have enough Flops to run the whole damn model. The network wouldn't support either option though.
I'm still not sure if this says more about Moore's law or the hamstringing of NWS's budget.
I'm still not sure if this says more about Moore's law or the hamstringing of NWS's budget.
That DNS not being integrated with Active Directory Domain Services can cause all sorts of problems. A client demoted a Domain Controller (properly-ish this time) and removed it form the domain. One of the other DCs ere giving a whole bunch of odd ball errors about being unable to replicate to a DC.
I checked that the meta data has been cleaned up, and thanks to them removing it correctly, it had. However the DC was one of the DNS name servers. The local tech then set up the new server using the same IP as the old server. DNS had 4 separate A records pointing to the same IP, and I could not delete, add, or edit any of them.
I puzzled around a bit, as the new primary name server could not edit the records either. I then discovered/remembered that the DNS records are still a flat file editable by notepad++. Edited the file, commenting out all instances of the old server, added the new name server. A refresh in the DNS tool and behold, I could add CNAMES and remove the remaining A record that didn't belong there.
I checked that the meta data has been cleaned up, and thanks to them removing it correctly, it had. However the DC was one of the DNS name servers. The local tech then set up the new server using the same IP as the old server. DNS had 4 separate A records pointing to the same IP, and I could not delete, add, or edit any of them.
I puzzled around a bit, as the new primary name server could not edit the records either. I then discovered/remembered that the DNS records are still a flat file editable by notepad++. Edited the file, commenting out all instances of the old server, added the new name server. A refresh in the DNS tool and behold, I could add CNAMES and remove the remaining A record that didn't belong there.
While it really depends on what extras arstechnica has wedged in, phpBB does have some naive handling of super-long threads.LordFrith":2m4lj2zz said:Since this is the server room, can we see something like a post-mortem on this? The Boardroom and Lounge have multiple threads of two to three times the size: what is different in this case? Is it a db partitioning scheme?continuum":2m4lj2zz said:Continuing the previous "what did you learn today", since that one appears to have gotten so large it's making the forum unhappy.
Learned how to change esxi default path selection policy to round robin for a particular SATP and apply to all LUNs after reboot.
Was a bit shocked to see active-passive and MRU on my new VNX..
Was a bit shocked to see active-passive and MRU on my new VNX..
...that while you can use the Unisphere VNX client to manage an EMC Celerra, you may wind up reverting to the built in client for various things, because it just won't appear to work right. Going back to the built in client and you can't do the initial steps very well either, so you start in the Unisphere VNX client and then revert to the built in client to do the actual iSCSI LUNs.
Ahhhh, more enterprisey crap.
In other news, would you believe that my job environment has taken a nearly 180 degree turn in the last few weeks, at least within my department?
And I have to say that it's all thanks to the guy who go the promotion by ass-kissing. Since he wants another promotion, that means he had to give up being the "phone guy" (literally only did voice mail and phone extensions on the Cisco VOIP setup we have) and actually take care of systems. Well, he's become more of a mouthpiece and works me hard to get him things to click and drop in Altiris, but he's actually been good at insulating me from little bits of busy work which keep him busy all day (should take 1/10th the time).
But, that means I'm spending less time on that. I build the pieces and set them in place, and let the others go and play.
And, it's working. The last piece was getting the dang DeepFreeze working properly. That took a bit of fighting, but it's working now.
Fixing the SQL DBs (reindex) that back the Altiris server and the vWorkspace was trivial enough to do and actually made things work right.
But, again I could only do that because that guy has managed to take enough heat off me that I can do my damn job.
So, NOW, I'm thinking maybe I'll stick around assuming I get the promised promotion (won't take effect until July 1st it seems due to funding cycle).
---
In other news, now that the Ex2010 install is finally released from all the blocks, I will now be hurrying up the Simpana v9 install (yes, v10 was released) to make sure it's supported for backup. I won't be migrating mailboxes until after the backup is working properly.
And even better, since the Ex2010 setup will be working, that means I can now set up an Email archive server to ingest all the durn PST files and then export the mailbox DB so our HQ org can ingest the result into the org-wide eDiscovery/Archiver product.
Ahhhh, more enterprisey crap.
In other news, would you believe that my job environment has taken a nearly 180 degree turn in the last few weeks, at least within my department?
And I have to say that it's all thanks to the guy who go the promotion by ass-kissing. Since he wants another promotion, that means he had to give up being the "phone guy" (literally only did voice mail and phone extensions on the Cisco VOIP setup we have) and actually take care of systems. Well, he's become more of a mouthpiece and works me hard to get him things to click and drop in Altiris, but he's actually been good at insulating me from little bits of busy work which keep him busy all day (should take 1/10th the time).
But, that means I'm spending less time on that. I build the pieces and set them in place, and let the others go and play.
And, it's working. The last piece was getting the dang DeepFreeze working properly. That took a bit of fighting, but it's working now.
Fixing the SQL DBs (reindex) that back the Altiris server and the vWorkspace was trivial enough to do and actually made things work right.
But, again I could only do that because that guy has managed to take enough heat off me that I can do my damn job.
So, NOW, I'm thinking maybe I'll stick around assuming I get the promised promotion (won't take effect until July 1st it seems due to funding cycle).
---
In other news, now that the Ex2010 install is finally released from all the blocks, I will now be hurrying up the Simpana v9 install (yes, v10 was released) to make sure it's supported for backup. I won't be migrating mailboxes until after the backup is working properly.
And even better, since the Ex2010 setup will be working, that means I can now set up an Email archive server to ingest all the durn PST files and then export the mailbox DB so our HQ org can ingest the result into the org-wide eDiscovery/Archiver product.
After all the FSM worthy pain of getting the balky storage working properly, the Ex2010 initial install has been cake.
Inbound and outbound internal works. Outbound external email also works. Inbound external email should start working after the Edge Subscription finishes f'ing around.
Inbound and outbound internal works. Outbound external email also works. Inbound external email should start working after the Edge Subscription finishes f'ing around.
What, a VMware product that doesn't require an additional expensive license? LIES!
shiznit":encvoj95 said:
Round robin pathing is default for VMW_SATP_ALUA_CX with vSphere 5.1. Are your LUNs properly configured for ALUA (failover mode 4)? Hosts properly registered in Unisphere?
ronelson":198s4d18 said:
Don't worry: it will use resources out of your existing VMWare license so you'll still pay them for using it so it's probably legit.
ronelson":jviq9nkb said:
There's actually a bunch of stuff like that, like projectknee
And they even contribute to openstack.
This was it, thanks! And the I/O analyzer is very cool, though for playback I think I like ioblazer better, the stupid analyzer only allows you to attach 1 vdisk per target for playback...Myrmidon":2wwfgbgd said:
Chargebacks of $2/GB/month? Granted, that does include some amount of backups. But is that at all within a reasonable range?
^^^ Depends on the tier of storage, the amount of support, data retention, SLAs (RTO/RPO), etc.
daldrich":2gtcwj9e said:
After doing research I discovered that apparently 2008r2 only partially supports UEFI PXE booting. And to use 2012 for full support.
euri":37n53f6a said:
Yeah, there are a ton of variables that come into play and there is no one right answer. Are you trying to subsidize some or all of the cost? Do you pay for power? Rackspace? What level of equipment? Backups?
sryan2k1":3p0yst55 said:
These are virtual machines inside the corporate data center. RTO/RPO are 72/24 hours I believe. And when they restore they will restore the DB but applications get re-installed on a clean system. From our perspective I don't really care what tier, but I believe they are giving us SAN. Not sure on retention.
Also, the VM itself (2cpu/4g) is something like 800/month.
We're also paying >$5k/mo for an HP DL380 with 64 GB and 1.5 TB storage with NO BACKUP OR SUPPORT (a test box)
I'm not coming here for ammunition to go argue with them. That is a lost cause. I'm more interested if this is normal in very large IT organizations (we're in the fortune 50). I heard from a source I don't completely trust that there are 'taxes' on all of the systems to fund other IT things and that some of the pricing is designed to discourage not following their approved patterns. I hate corporate politics.
That this http://www.cjwdev.co.uk/Software/ADPhotoEdit/Info.html looks like a great way to automate the injected of staff photos in AD.
That seems insanely expensive.
We're also paying >$5k/mo for an HP DL380 with 64 GB and 1.5 TB storage with NO BACKUP OR SUPPORT (a test box)
So does that.
Google compute engine/AWS instances of that size are around $170-190 per month, I thought that was insanely expensive compared to what I can provide an instance for internally but apparently they're competing against a whole different set of economics.sryan2k1":ywhf2fhp said:
Pretty sure that is almost enough to purchase that HP server over and over each month...sryan2k1":17jqf8fq said:
That seems insanely expensive.
We're also paying >$5k/mo for an HP DL380 with 64 GB and 1.5 TB storage with NO BACKUP OR SUPPORT (a test box)
So does that.
At that level of chargeback, especially with the apparently poor level of support, I would just run it on an outside service that has reasonable security expectations like RackSpace or similar. Use the excess budget to get a second of everything or overspec.
Yeah, our VMWare hosts 2 x 10 Core (some E7-xxxx) and 512GB RAM with 10Gb NICs and no local storage costs us ~30k. You could buy one of those in 6 months.
And I thought the $45k cost of each blade (complete cost, CAPEX + OPEX x 3 years) that we dealt with in my former life was bad...
Pretty close to our rates, not sure if you can see this page without login:
http://www.stanford.edu/services/itrates/computing.html
I'm not saying it's not insanely high, but par for the course for large bureacracy-laden organizations.
I found your problem...
I am guessing another problem is that these kind of ridiculous chargebacks were a holdover from some other inter-departmental chargeback system that's unrelated to virtualization, where both sides had a large pool of funny money so if you divided the charges by ten or multiplied by ten, the pools would shrink or grow to match. Then someone used the same scaling for something that involves real money. For instance, internally we get charged $400 if we have to call the helpdesk to reset our domain password. No money ever changes hands, but some report shows that our org spent $1600 this month on password resets and the password reset org made $1600 from us and a total of $2 hojillion (we have stupid users, I'm not sure that's an exaggeration!). That's fine. The password reset group shows their "value" and we get our passwords changed. But if we actually had to fork over $400 in cash, things would be a LOT different.
Rick25":3l3cwjhx said:
Very nice timing, I was looking into doing this yesterday.
Duckie Dooh":pyfgcccz said:
My guy ended up using http://www.exclaimer.com/outlook-photos/ when we put ours in. Worked really well and free.
Duckie Dooh":1u388yjd said:
Heh. Make sure you get buy in from senior management. Our Sharepoint team pulled me in to assist with PowerShell for this; pull data from badge database, resize and/or compress images, check for overrides (e.g. certain folks have 'professional' pictures to use instead of badges), update AD. Asked if they had discussed with HR / management, if and how they would handle exceptions...
Everything was working great on the technical side, our department all had images for testing purposes.
Never underestimate the push back you will get on displaying photos of other people. Even if they are legally obligated to wear a badge with the same picture on it. Sadly, the idea was scrapped when the complaints started pouring in.
It seems like a small thing, but having the pictures in Lync, Outlook and Sharepoint was quite handy, I miss it.
One of our apps guys wrote a thing to pull images from our contact management system and stuff them into AD. It's nice to see pictures finally show up in Lync/Outlook.
speedye1":1b8dc8pu said:
Free is great, and it looks like it is easier to use then ADPhotoEdit. Powershell would have been fine if the pictures were standard or even uniform sizes.
We do have management buy in, they are the ones that miss it after we moved from on-prep to Office 365. Now if only the "Microsoft Back End Team" could get back to me on why address lookups in Lync are not working...
Thanks all!
TIL that when you install server manager on Windows 8 UK English, it doesn't.
You actually have to install the US English language pack, and then re-run the install before it works
You actually have to install the US English language pack, and then re-run the install before it works
daldrich":gc65zdyl said:
Yup, my favourite is where it'll lay down a UEFI partition scheme for any image, even if the image was captured from a legacy machine (and hence doesn't support booting off a UEFI boot partition)... diskpart to the rescue
Batz_10K":2jeofyoq said:daldrich":2jeofyoq said:
Yup, my favourite is where it'll lay down a UEFI partition scheme for any image, even if the image was captured from a legacy machine (and hence doesn't support booting off a UEFI boot partition)... diskpart to the rescue
Yup. Found another quirk, which could have just been with that server. But even upgrading the 2008r2 box to 2012 still wouldn't let it work.
Wonder if I can put this on my resume. Can image UEFI machines
TIL just how much I need internet. Yesterday, Windows informed me that it would take 2.5 days to download a 254M file. What the...? FIOS is now giving me 200kbps down, and a whopping 30Mbps up. And cuts out intermittently, which means that file kept failing at <10 meg anyway. Apparently this is some issue that may affect more than 1000 customers so they won't do anything about it until 1AM. Fuck you, Verizon! So now I can't work, can't play online games, and can barely even get to Ars. And I have to keep calling them to escalate (they've told me three times now that it will be fixed in a 1AM maint window, so I'm not convinced). What a boring, depressing day.
On the other hand, all of the people *at* Verizon have been pleasant to work with. It's some fuckers in the CO who don't talk to the outside world who I despise right now.
On the other hand, all of the people *at* Verizon have been pleasant to work with. It's some fuckers in the CO who don't talk to the outside world who I despise right now.
What, no tethering? That's what I do when I have to connect to work and the cable's down (our maintenance window has coincided with my cable companies more than once).ronelson":2601qqya said:
I have a grandfathered unlimited plan, and I'd like to keep it that way. Plus, uh, it's work
Looks like they keep "fixing" it to be more better:
I learned that in 2014 my company is moving us to Vancouver.
No, not that Vancouver - land of good times, excellent food, attractive and sophisticated people, etc. etc. This Vancouver.
GODDAMMIT!
My 30-minute walking commute will change to a 45-minute driving commute. Which means I'll have to buy a car. Which means I'll have to come up with several thousand dollars up front and my monthly, after tax outgoings will increase by $200-300.
GODDAMMIT AGAIN!
No, not that Vancouver - land of good times, excellent food, attractive and sophisticated people, etc. etc. This Vancouver.
GODDAMMIT!
My 30-minute walking commute will change to a 45-minute driving commute. Which means I'll have to buy a car. Which means I'll have to come up with several thousand dollars up front and my monthly, after tax outgoings will increase by $200-300.
GODDAMMIT AGAIN!
FTFY, because it's not like all viruses let you know they're there. I'm betting you've had Java trojan shit camping out on your systems for years. Probably a keylogger or two as well.gradster":3nps1r25 said:
General rule of thumb: if you have no malware protection and you don't think you've been hit by malware...you've been hit by malware.
hux":1qaais4q said:
Eh, I ran without AV for 3 years or more, up until about 2008 or so. It's definitely possible to do it if you know what you're doing, and what you use the machine for.
That said, at this point there's really no purpose to not running MSSE, and it's built-in to newer OSes...so yeah.