|
spoon daddy posted:Depends on cooling but in general, they aren't quiet. Hmm, well... It's the loudest goddamn thing I've ever heard in a data center. If I stand behind it it blows my hair around. It is being used very heavily (two 10GbE links I think, and it runs a database that has extremely heavy traffic), but still. This is the appliance itself I'm talking about, not the shelves.
|
#
?
Jun 20, 2012 23:23
|
|
|
|
| # ? Apr 19, 2024 06:48 |
|
|
Equipment that's being used at the top of its performance envelop will typically also operate at the top of its thermal envelop. And thus generate a metric shitton of heat, which is typically removed in the form of hot air by very noisy fans.
|
|
#
?
Jun 21, 2012 00:11
|
|
|
Internet Explorer posted:Which really, in my opinion, is not so much the fault of the admin and more the fault of management. It is their responsibility to make sure procedures are in place to enforce these types of things. This is true and problems like the one PTM is having are large scale organizational failures so it's not entirely fair to blame one dude for all of it. But I spent a couple of years as a lead for a group of sysadmins and despite providing examples of what I expected I don't think I ever received an acceptable project plan from any of them. This isn't to say that they were bad sysadmins or stupid, most weren't, but they had never learned anything about the organizational aspects of managing a large or critical environment. They viewed their job as basically an extension of tinkering on their android phone or home linux box. If it was possible to accomplish something technically that was enough license to proceed with doing it, with no further thought or planning. I got a lot of that "measure twice, cut once" mentality drilled into me while working as a consultant where it's a lot more important that you emphatically do not gently caress up the customer's stuff and you have a crapload of documentation to support any decision you make, lest it become a reason for the customer not to pay you.
|
|
#
?
Jun 21, 2012 00:54
|
|
|
I just got off the phone with the Rep from Dot Hill along with a rep from EMC (Zones conferenced) and here is what I took away: EMC supports NFS and iSCSI, while Dot Hill is only iSCSI. (I think we are going to go with iSCSI for VMWare). The chassis for the VNXe 3300 is only 3.5 bay while I can start out the Dot Hill with 24 2.5 in a 2U package. Drives need to be bought directly from both companies, but with the Dot Hill I can buy say 1 at a time instead of a pre-configured drive pack. The Dot Hill comes with 2 active/active controllers with 4 network ports each. That means 8 usable ports. Dot Hill's RAID configuration is stored on volume, which means that I can upgrade controllers without migration. Software is the same it seems except app specific backups like Exchange and SQL. Although I will be using Veeam to backup in any case. The whole package from Dot Hill running ~$7k less. Dot Hill has remote installation service and support that is cheaper and includes full documentation along with post install support. Dot Hill is also ~$12k cheaper it seems. I looked into the VNXe 3100, but I do not see much difference other than some software differences. Next up on the quote carousel is the Dell solution.
|
|
#
?
Jun 21, 2012 01:04
|
|
|
If all you need is iSCSI do yourself a favor and take a good hard look at the Dell/Equallogic SANs. EMC is such a pain in the rear end.
|
|
#
?
Jun 21, 2012 01:23
|
|
|
NippleFloss posted:This is true and problems like the one PTM is having are large scale organizational failures so it's not entirely fair to blame one dude for all of it. But I spent a couple of years as a lead for a group of sysadmins and despite providing examples of what I expected I don't think I ever received an acceptable project plan from any of them. He has a book about ITIL on his desk. The irony is hysterical. Edit: Sorry for the derail, I feel like I'm talking about drama now. I'll stop.
|
|
#
?
Jun 21, 2012 01:32
|
|
|
somecallmetim posted:Quote stuff Netapp has a FAS2240-2 that uses 24 2.5 drives in a 2u enclosure and does FC, iSCSI, NFS and CIFS. Also supports ssd cache layer, external shelves 10gbe and some other stuff you might not find on arrays at this price point.
|
|
#
?
Jun 21, 2012 01:42
|
|
|
somecallmetim posted:I just got off the phone with the Rep from Dot Hill along with a rep from EMC (Zones conferenced) and here is what I took away: It sounds like the Dot Hill system you are describing is what HP is reselling as the P2000 G3, so you may want to cross shop there as well. http://h10010.www1.hp.com/wwpc/us/en/sm/WF25a/12169-304616-241493-241493-241493-4118559.html?dnr=1
|
|
#
?
Jun 21, 2012 03:07
|
|
|
spoon daddy posted:Depends on cooling but in general, they aren't quiet. There's also a chance you've hit a bug. You should run environment status from either the RLM/SP/BMC or straight SSH (depends on your DataONTAP version what output you will get) and look at the fan RPMs. If they are reading ~6000 that's the max speed and I would question whether they need to be that. It depends on your ambient temperature etc but in a typical install even a fully thrashed controller will sit at ~3000 RPM. You can fix this by popping off the front face plate, pulling one of the fan modules out, waiting a moment, putting it back in. Done.
|
|
#
?
Jun 21, 2012 05:06
|
|
|
Powdered Toast Man posted:He has a book about ITIL on his desk. The irony is hysterical. No no, please don't stop. This is an SHSC soap opera in progress, don't blue ball us.
|
|
#
?
Jun 21, 2012 05:16
|
|
|
Powdered Toast Man posted:He has a book about ITIL on his desk. The irony is hysterical.
|
|
#
?
Jun 21, 2012 05:51
|
|
|
So I verified that the appliance in question is a Reldata 9400 12TB. Does anyone know about the performance of such a system (note in particular that I'm pretty sure it has 7200rpm drives) when used as a CIFS NAS? They are gradually getting the permissions issues fixed but as that happens the usage level of the appliance is also increasing and I'm wondering how long it will take before it self-destructs from the strain.
|
|
#
?
Jun 21, 2012 15:06
|
|
|
Powdered Toast Man posted:So I verified that the appliance in question is a Reldata 9400 12TB. Does anyone know about the performance of such a system (note in particular that I'm pretty sure it has 7200rpm drives) when used as a CIFS NAS? They are gradually getting the permissions issues fixed but as that happens the usage level of the appliance is also increasing and I'm wondering how long it will take before it self-destructs from the strain. If it's like the Reldata I evaluated it's doing the following: *Using three disks at least for OS and hot spare. Which of the disks are those? No idea! *One side is active with no real load balancing, which means all your traffic is going out the same 1GB NIC *None of the email notifications are because you changed the default admin password I see all you guys complaining about EMC but at least they aren't these rebadged supermicro shitheel systems that I deal with. I would give my nuts for the worst in class EMC system at this point 
|
|
#
?
Jun 21, 2012 16:24
|
|
|
Netapp just delivered my 10ksas tier. It's full of 15k drives. Do I call them?
|
|
#
?
Jun 21, 2012 16:29
|
|
|
evil_bunnY posted:Netapp just delivered my 10ksas tier. It's full of 15k drives. Do I call them? I would at least check the invoice vs packing slip to make sure the serial numbers match. And yes, calling your rep is a good idea, because they can refuse warranty on the drives (or the entire unit, really) if it turns out you got the wrong ones. Since the drives are the number 1 most likely item to need a warranty... Sucks if they want them back, but you can make them do any switch on their time and their dime, because a mis-ship isn't your problem to fix.
|
|
#
?
Jun 21, 2012 16:42
|
|
|
I haven't really had a chance to take a look at this myself yet, but any tips/links/advice for someone who's been supporting Cisco MDS/Nexus switches and will soon need to jump into management of Brocade? We'll have some test gear to play around with and will be able to lean on Brocade for some training and such, but any other helpful stuff would be awesome!
|
|
#
?
Jun 21, 2012 16:46
|
|
|
EoRaptor posted:Sucks if they want them back, but you can make them do any switch on their time and their dime, because a mis-ship isn't your problem to fix.
|
|
#
?
Jun 21, 2012 16:49
|
|
|
qutius posted:I haven't really had a chance to take a look at this myself yet, but any tips/links/advice for someone who's been supporting Cisco MDS/Nexus switches and will soon need to jump into management of Brocade?
|
|
#
?
Jun 21, 2012 17:30
|
|
|
Just FYI, Brocade's Data Center Fabric Manager (generally abbreviated DCFM) is being phased out. FC and IP management are being converged under Network Advisor.
|
|
#
?
Jun 21, 2012 17:37
|
|
|
complex posted:Just FYI, Brocade's Data Center Fabric Manager (generally abbreviated DCFM) is being phased out. FC and IP management are being converged under Network Advisor. 
|
|
#
?
Jun 21, 2012 18:32
|
|
|
evil_bunnY posted:Netapp just delivered my 10ksas tier. It's full of 15k drives. Do I call them? 10k SAS drives are only offered in the 2.5" form factor, so if they are 3.5" disks they have to be 15k. If they are 2.5" then they can only be 10k as we don't have any disks that support 15k at that size. YOLOsubmarine fucked around with this message at 19:14 on Jun 21, 2012 |
|
#
?
Jun 21, 2012 18:58
|
|
|
Powdered Toast Man posted:Uh...yeah, about that. We actually have a Netapp FAS3420 with two shelves. It's being used for our "really" mission critical stuff, but originally that stuff was on Reldata hardware as well. All the PSUs are getting power right? I was staging a FAS2040 in a lab that was only half plugged in and it was exceptionally loud until I was able to get the other PSUs plugged in.
|
|
#
?
Jun 21, 2012 19:00
|
|
|
NippleFloss posted:10k SAS drives are only offered in the 2.5" form factor, so if they are 3.5" disks they have to be 15k. If they are 2.5" then they can only be 10k as we don't have any disks that support 15k at that size. In any case that's good to know, thanks.
|
|
#
?
Jun 21, 2012 19:19
|
|
|
Dumbass IT Director: "Yeah I think we're gonna roll back and abandon the Reldata..." Starboard Support: "NO WAIT WE HAVE A HOTFIX" Dumbass IT Director: "A HOTFIX YOU SAY WELL I NEVER" Briefly, my life was looking up. A ray of sunlight was visible, as it appeared they realized they needed to jettison the Reldata and get a real filer. I then heard, as from very far away, a sound that was very much like a giant, $700 million toilet flushing. Previous to this Starboard kept going "DAFUQ" and had no idea what the problem was. Apparently something is going on with communication between the NAS and active directory when ACLs are interrogated, I don't fully understand it. What it really boils down to is load, and this filer cannot handle the load that is being put on it. Even if they get it "working" it's going to eventually turn into a very expensive pool of molten metal on the data center floor. I hate my life. 
|
|
#
?
Jun 22, 2012 01:04
|
|
|
|
|
#
?
Jun 22, 2012 01:26
|
|
|
ghostinmyshell posted:If it's like the Reldata I evaluated it's doing the following: What's wrong with custom supermicro gear?  (420tb, 12 cores, 196gb ram, SSD's, and Infiniband running NexantaCore and Napp-IT.) Our systems are essentially giant WD-Mybooks though
|
|
#
?
Jun 22, 2012 03:43
|
|
|
the spyder posted:(420tb, 12 cores, 196gb ram, SSD's, and Infiniband running NexantaCore and Napp-IT.) Are these designed to be portable sans or something? What are you doing that you need that much data to be portable (if they are indeed portable)?
|
|
#
?
Jun 22, 2012 04:28
|
|
|
Goon Matchmaker posted:Are these designed to be portable sans or something? Those are SKB cases used for shipping servers and in our case overseas. These are archival storage for images and that is about all I should (read:NDA) say. I can say that we have a few dozen of these in various size configs with little to no issues in 100+ degree environments.
|
|
#
?
Jun 22, 2012 04:55
|
|
|
the spyder posted:Those are SKB cases used for shipping servers and in our case overseas. These are archival storage for images and that is about all I should (read:NDA) say. I can say that we have a few dozen of these in various size configs with little to no issues in 100+ degree environments.
|
|
#
?
Jun 22, 2012 12:00
|
|
|
You know, I think if your job is so stressful and so much pressure is being put on you that you start to ponder , that is a bad sign. That Careerbuilder commercial is on repeat in my mind.
|
|
#
?
Jun 22, 2012 17:53
|
|
|
Powdered Toast Man posted:You know, I think if your job is so stressful and so much pressure is being put on you that you start to ponder Take some time off. No job is worth sacrificing your mental health for. 100% serious, do it.
|
|
#
?
Jun 22, 2012 18:43
|
|
|
I have 8x 1.5TB drives in a JBOD enclosure with dual sas ports and a LSI RAID card to go with it. I'm considering using a not so old machine I have laying around with a dual core proc and 8 gigs of ram and a smallish SSD and doing ZFS for file sharing at home. I currently have an n40L for Home Server 2011 but would like to offload the storage of music and video on an actual file server / NAS. I've seen a lot of talk of ZFS and just wonder what the preferred implementation is. FreeNAS, Nexenta, Open Solaris? I've rolled my own NAS in the past with Linux and NFS / iSCSI but want to try something different. Ideally I can get it running and just forget about it. I'd like notifications on drive failure or SMART alerts. What do you recommend?
|
|
#
?
Jun 22, 2012 21:58
|
|
|
OpenSolaris is long dead, you're looking at OpenIndiana at this point. I have absolutely zero confidence in the ability of the open source standard-bearers to continue the project successfully, so I'd count them out in favor of one of the other options. We recently settled on vanilla FreeBSD 9 for our environment. We're running a bit bleeding-edge with Samba 3.6, but it handles ZFS's NFSv4 (CIFS) ACLs quite nicely.
|
|
#
?
Jun 22, 2012 22:01
|
|
|
I had a director at my desk today asking why I haven't started migrating to the 100TB of Hitachi he'd bought to replace our ageing 3PAR frames. I asked him where the management tooling was to allow us to troubleshoot performance issues. His response was that the American office had been using it for two months. I responded by opening an email about their VSP outage that's still open between VMware and Hitachi. He advised he was going to yell at engineering about getting the management suite setup. Gold.
|
|
#
?
Jun 22, 2012 23:28
|
|
|
Misogynist posted:OpenSolaris is long dead, you're looking at OpenIndiana at this point. I have absolutely zero confidence in the ability of the open source standard-bearers to continue the project successfully, so I'd count them out in favor of one of the other options. I have a demo copy of TrueNas with Infiniband support I am going to start testing here next month. In house the devs are claiming 3.2GB's on 40Gb Infiniband. Not bad.
|
|
#
?
Jun 23, 2012 04:06
|
|
|
evil_bunnY posted:They must have delivered a -4 instead of a -2 I guess? I haven't opened that box yet, but the hot spare that was delivered separately is definitely 3,5"
|
|
#
?
Jun 25, 2012 15:22
|
|
|
What's that you say? ANOTHER hosed UP SHARE?  His boss is back in town tomorrow. I can only hope...
|
|
#
?
Jun 25, 2012 17:15
|
|
|
We're migrating exchange from 2k3 to 2k10. They decided to move 2k3 to the new Netapp array before doing the upgrade. 336 15k disks in the old exchange storage, 72 in the new. What could possibly go wrong?
|
|
#
?
Jun 25, 2012 18:53
|
|
|
Nomex posted:We're migrating exchange from 2k3 to 2k10. They decided to move 2k3 to the new Netapp array before doing the upgrade. 336 15k disks in the old exchange storage, 72 in the new. What could possibly go wrong?  at trying to run 2003 on that
|
|
#
?
Jun 25, 2012 18:57
|
|
|
|
| # ? Apr 19, 2024 06:48 |
|
|
Nomex posted:We're migrating exchange from 2k3 to 2k10. They decided to move 2k3 to the new Netapp array before doing the upgrade. 336 15k disks in the old exchange storage, 72 in the new. What could possibly go wrong? How big of an Exchange environment? 72x 15k disks could handle about 3,000 IOPS.
|
|
#
?
Jun 25, 2012 21:26
|
|
