SCVMM is really shitty compared to vCenter. No, I didn't learn this today, but it was reinforced.
what did you learn today?
- Thread starter chalex
- Start date
- Status
I just learned why HP went with RAID6 on their VLS9000 series...4 hard drive failures in just under an hour. 2 different enclosures, so 2 different RAID groups. No data loss yet (knock on wood). This unit has been in for about 2 years and has had the random drive fail, but this is the first time I've ever had 4 drives die within an hour of each other on the same unit. 2 of the drives are on a shelf that is half a year old...what a fun thing to walk into first thing in the morning.
hutchingsp":2kwo193s said:
I know with EVA we just stop the DESTA service on the CV:EVA node and it ceases all email notifications until we get done patching and start it up.
Many developers really are clueless.
From: Random developer dude
To: Every possible IT distribution list
SUBJ: ports not working on (hostname)
Hi network team, below ports are not working on (hostname)
@hostname:/root
=>telnet localhost xxxx
trying 127.0.0.1...
connect to address 127.0.0.1: connection refused
trying ::1
telnet: unable to connect to remote host, network is unreachable
(repeat above for several more port numbers)
Okay, so perhaps somebody needs to update that stupid old error text...but really, does a developer not understand that requests to localhost will not traverse the network?
From: Random developer dude
To: Every possible IT distribution list
SUBJ: ports not working on (hostname)
Hi network team, below ports are not working on (hostname)
@hostname:/root
=>telnet localhost xxxx
trying 127.0.0.1...
connect to address 127.0.0.1: connection refused
trying ::1
telnet: unable to connect to remote host, network is unreachable
(repeat above for several more port numbers)
Okay, so perhaps somebody needs to update that stupid old error text...but really, does a developer not understand that requests to localhost will not traverse the network?
Today I learned that due to "Access Based Enumeration" you can replicate Netware file shares fairly well on Windows 2008R2. Add in trustee.nlm to dump the permissions on the Netware box and two fairly gnarly Python scripts, you can even "migrate" user permissions.
That certain nvidia* SATA chipsets (MCP55 rev a3) have a bug that cause ATA bus errors on kernels 2.6.32 or greater.
*yes, I know, fucking nvidia get-yourself-a-real-chipset-sonny-jim. It's a cheap home test box (Proliant ML115).
*wanders off in the vain home HP will have released a firmware patch*
*yes, I know, fucking nvidia get-yourself-a-real-chipset-sonny-jim. It's a cheap home test box (Proliant ML115).
*wanders off in the vain home HP will have released a firmware patch*
gchapman":2n0vpjev said:
Which version of CallXpress and what feature\capability did they claim? (Just curious)
Its 8.11 and FYI Lotus integration is pretty much not all that they claim.Widger":20o94erx said:
Oh and an "engineer" stated that they had never seen a 9GB email file before which was good for a chuckle, I can't wait to see what he says when he looks at the CEO's 35GB mailbox.
Yeah, our lead counsel had a 42GB archive and a 20GB active mailbox when we moved off Notes. The only reason the 42GB file was an archive is that the full text indexing stopped working.gchapman":26c7ut1c said:
...that I can't upgrade directly from a FortiGate-60 to a FortiGate-60C, because 60 doesn't run FortiOS 4, and 60C doesn't run FortiOS 3. Fortunately I had a spare 60B on a shelf, which I downgraded to 3.7.9, transplanted the config from 60, upgraded to 4.1.6, then to 4.2.6, and then transplanted the config into the new 60C. Since the interface layout on all three is the same (internal/wan1/wan2/dmz/modem/ssl.root) I only had to swap headers.
I have a feeling I will be repeating this particular dance in the coming months, as we've got 11 more 60s on supported customer sites, and I hear Fortinet is going to completely EOL (as in, no more support contracts) this model fairly soon.
I have a feeling I will be repeating this particular dance in the coming months, as we've got 11 more 60s on supported customer sites, and I hear Fortinet is going to completely EOL (as in, no more support contracts) this model fairly soon.
I managed IT at a boutique BK firm during Enron and we were involved and routinely received 10,000 emails a day on just that case alone, oh and we were running Exchange 5.5 at the time. I've worked at several law firms over the years and they have a penchant for huge email files and archives. I'm sure things have gotten better from an archiving standpoint since I last worked at a firm, but 40 and 50 gig mailboxes were common for certain types of litigation, especially IP and Patent.afidel":3glgxqps said:
gchapman":1tni3lc1 said:
Today, though, don't those types of companies start playing with email vaulting and/or other tech that both
1) aid in legal discovery, document retention (and thus deletion!)
2) keep actual mailbox sizes down
?
you would think so wouldn't you. then again, in most law firms IT is run by a committee of people with little to no real technology background, the person who was capable enough to synch their blackberry on their PC. sure its not like that at all firms, but the 3 I worked for it was. decisions are made slowly, and given that most capital investment comes out of the partners pockets, things tend to be done on the cheap unless their can be a very quick ROI or they can bill its use to a client.finni":mcghiw78 said:
Bahahaha. NetApp will sell you a DS14mk2 with 14x300GB for $12.5k, or a FAS2020, dual controller, 12x600GB, all licenses included, 3 years of maintenance, for $13k. Considering that we initially requested a quote for a DS14mk2 to add capacity to a two-year-old FAS2020, it just makes no sense not to get the full system.
Don't you lose a significant amount of space to the OS though?Barmaglot":244q31m5 said:
You do, but with almost double the drive space (12x600GB vs 14x300GB) you still come out ahead.
Edit: After all overhead, and allowing for 10% aggregate free space, a 300GB data drive can store 213GiB; a 600GB correspondingly has 426GiB. If we add 14x300GB to the current FAS2020, which is split 10/2 (first controller has 10 drives in RAID-DP - 7 data, 2 parity, 1 spare; second controller has 2 drives in RAID4 - 1 data, 1 parity, no spare), then we'll be able to give each controller 13 drives, for 10 data drives, 2 parity and 1 spare for each controller, adding an effective 2556GiB (12x213GiB) of capacity to the system. With a new, 12x600GB FAS2020, in a 10/2 split, we'll get 3408GiB, or 2982GiB if we ignore the second controller with 1 data drive and no spares. In the least-efficient 6/6 split, where each controller has 3 data drives, 2 parity and 1 spare, we still get the same 2556GiB of added capacity, while also getting a new maintenance contract (rather than paying to renew maintenance on the old FAS2020 in one more year), all the licenses (flexclone, SMVI, SMSQL, Exchange including single mailbox recovery, etc), and extra controllers to spread the load around, which is not unimportant with the weak FAS2020 CPU and RAM.
Edit: After all overhead, and allowing for 10% aggregate free space, a 300GB data drive can store 213GiB; a 600GB correspondingly has 426GiB. If we add 14x300GB to the current FAS2020, which is split 10/2 (first controller has 10 drives in RAID-DP - 7 data, 2 parity, 1 spare; second controller has 2 drives in RAID4 - 1 data, 1 parity, no spare), then we'll be able to give each controller 13 drives, for 10 data drives, 2 parity and 1 spare for each controller, adding an effective 2556GiB (12x213GiB) of capacity to the system. With a new, 12x600GB FAS2020, in a 10/2 split, we'll get 3408GiB, or 2982GiB if we ignore the second controller with 1 data drive and no spares. In the least-efficient 6/6 split, where each controller has 3 data drives, 2 parity and 1 spare, we still get the same 2556GiB of added capacity, while also getting a new maintenance contract (rather than paying to renew maintenance on the old FAS2020 in one more year), all the licenses (flexclone, SMVI, SMSQL, Exchange including single mailbox recovery, etc), and extra controllers to spread the load around, which is not unimportant with the weak FAS2020 CPU and RAM.
That working in an evironment where Domain Users are local Administrators is crazy.
dp2":1dosy37q said:
Couldn't agree more. Especially when the primary software used requires them to be local administrators.
Use Process Monitor to find out where the app is trying to write to and give access to the necessary registry keys and file locations and it should work fine.daldrich":1rnkjz87 said:
daldrich":302qip2p said:
This gets even more fun where some local users are administrators & there's no record of who's been given that right (configured locally, not through GP), ideal when you've been dropped in to do a domain / desktop migration...
Incarnate":okx00wah said:
We tried that and they were getting weird/funky errors so we just put it back.
I'm in this boat. Even though the software is better and probably doesn't require local admin rights anymore, just try to take this away. If we did this, we'd have a 'cultural revolution' to deal with.daldrich":hwjzoaig said:
And one more thing, please for the love of all that is good and holy, please gadgetwhore CEO, stop telling everyone how cool your new iPad is! It does not make my life easier.
...that I'm about ready to install a keylogger on one of my servers, the one that hosts the ADP program. This way, the next time a punch clock is replaced, and ADP had to Bomgar into the machine to give a password to restore the database to the clock, that I'll capture that password (which they won't give out, or course), and thus I can do almost all ADP troubleshooting myself. Support from them is like support from Symantec.
Just be careful that doing that doesn't violate your contract with ADP. Pissing off your HR and/or Payroll people because your endangered their relationship with a very important vendor isn't a Done Thing.Digitlman":3gj49fa8 said:
Digitlman":ryjlr4dz said:
DO NOT RECOMMEND.
Installing anything like that on a server that processes employee payroll data (even if it's just timeclock punches) is like asking to be fired.
Your dissatisfaction with a vendor's support does not trump the company's data integrity.
dp2":2wl43s3a said:
I don't find it that bad, as long as you have a good deployment scheme set up.
When people fuck up their computers due to being local admin, or complain about it being slow, we just hand them a new one with data migrated. Of course, only the business apps are installed, so they end up taking time to install all the non-essential junk themselves. If their manager asks why user X has been spending so much time "getting their computer back the way they like it", we're honest with the manager.
And it's great not to have to field questions like "can you please install this software", especially when the software requested is not required for their job description.
Digitlman":87qpyxl3 said:
You've been in IT too long when you can't dream anymore.
I am unfortunately, very keenly aware. Try one where someone put Authenticated Users into Enterprise Admins............dp2":2n8vig56 said:
daldrich":34mzfi7q said:
Yup, BTDT with various apps. The worst culprit is a federal student aid database package. It's based on Access, but has other crap added that make Power User the minimum requirement. I've got other various legacy apps that are slowly being replaced, but that sort of issue is still there for a number of them (lucky very small in number).
Digitlman":34mzfi7q said:
Yup, that's all it is, because you wouldn't have posted that you were going to do it, if you were really going to do it.
---
I know the feeling. I've gotten most of the credentials I need to get into our systems. It turns out that the SAN guy, VMWare guy and Data Domain/Commvault guy at the main contractor were hired out by EMC. I'm sure that it at least doubled their pay, given how this company underpays their staff.
Their staff turnover on this project alone has been close to 100% over the 1.5 years of the project thus far.
So, the reason the consultancy didn't give up the passwords and provide the training, is that they didn't have those internal resources any more.
I feel like quotes today.
"Build it and they will come", unless it is TSM and it will come and rape your network no matter how big the pipes are.
"Good judgment comes from experience. Experience comes from bad judgment" Apply to your IT woes as necessary.
"Build it and they will come", unless it is TSM and it will come and rape your network no matter how big the pipes are.
"Good judgment comes from experience. Experience comes from bad judgment" Apply to your IT woes as necessary.
A small pile of 100mit fiber to copper media convertors that were declared "broken" by our highly paid networking contractor, were not in fact not broken. The female tight leather pants CCIE had been connecting fiber runs with gigE fiber on the switches with said convertors and wondering why they didn't work.
If only she had read the words on the convertors, she would have quickly recognized the issue, instead of spending a day of troubleshooting one farkin fiber run and attempting to repatch everything. Of course, while doing this, my coworker was ordered to assist her, which mostly consisted of her putting her leather clad butt in front of his face while unnecessarily being ordered to steady her ladder. Pleasant for him, I suppose.
This happened a year ago.
The good news is, that their incompetence has lead them to running late on the project, but without getting additional money. IOW, for the last 6 months, they've been losing money because they haven't completed the job. Their latest CCIE is not an internal hire, but a subcontractor, for whom they are paying approximately $600 per hour (according to my sources).
Thank goodness, it doesn't look like they're going to get that support contract that they hoped for. Heh.
EDIT: we're reusing the fiber convertors as well as an old Cisco 2924XL for our building management system network. Works out well that way. We're tossing the 10Mbit fiber convertors, except where explicitly required by legacy crap.
----------
What's more worthwhile? VMWare classroom course that leads directly to VCP exam or EMC SAN coursework? Both are applicable to my career. (I'll be doing a MCITP exam later this month and CCNA probably next month but maybe this month, with only the CCNA costing me money)
If only she had read the words on the convertors, she would have quickly recognized the issue, instead of spending a day of troubleshooting one farkin fiber run and attempting to repatch everything. Of course, while doing this, my coworker was ordered to assist her, which mostly consisted of her putting her leather clad butt in front of his face while unnecessarily being ordered to steady her ladder. Pleasant for him, I suppose.
This happened a year ago.
The good news is, that their incompetence has lead them to running late on the project, but without getting additional money. IOW, for the last 6 months, they've been losing money because they haven't completed the job. Their latest CCIE is not an internal hire, but a subcontractor, for whom they are paying approximately $600 per hour (according to my sources).
Thank goodness, it doesn't look like they're going to get that support contract that they hoped for. Heh.
EDIT: we're reusing the fiber convertors as well as an old Cisco 2924XL for our building management system network. Works out well that way. We're tossing the 10Mbit fiber convertors, except where explicitly required by legacy crap.
----------
What's more worthwhile? VMWare classroom course that leads directly to VCP exam or EMC SAN coursework? Both are applicable to my career. (I'll be doing a MCITP exam later this month and CCNA probably next month but maybe this month, with only the CCNA costing me money)
Why are they paying ANYONE $600 to connect fiber convertors? We use NCR or CO techs for all our wiring needs, the CCIE's are not allowed to touch any of that (one good thing about the unions).
- Status