|
Nope, you're fine to not have a spare, and you can manually reassign the spare to either head at any time since you're using software disk ownership. Run with 1 spare, and use "disk assign -f" to manually move the spare to whichever head has a disk fail first.
|
#
?
Mar 18, 2013 19:39
|
|
|
|
| # ? Apr 26, 2024 04:26 |
|
|
madsushi posted:Nope, you're fine to not have a spare, and you can manually reassign the spare to either head at any time since you're using software disk ownership. Run with 1 spare, and use "disk assign -f" to manually move the spare to whichever head has a disk fail first. You're fine to not have a spare on controller 2, but it will complain about not meeting the minimum spares requirement and there is no way around that. It expects at least 1 spare on a controller with a raid-4 aggregate, and will not let you tell it otherwise. However that's purely cosmetic, so if you're fine with that complaint then go for it. The controller will not shut down based on a low spares condition but it WILL shut down if you have to run in degraded mode (a single disk failure in a raid-4 aggregate or a double disk failure in a raid-DP aggregate) for more than 24 hours. You can disable that behavior with some option settings, though it isn't recommended. You can run a 9 disk aggregate, plus 1 spare, on controller 1, and a 2 disk raid-dp aggregate on controller 2 with no spares and that's the safest way to get the maximum amount of disk on controller 1. As Madsushi said, if you need the spare on controller 2 due to a disk failure you can always move it between controllers.
|
|
#
?
Mar 18, 2013 19:52
|
|
|
NippleFloss posted:and a 2 disk raid-dp aggregate on controller 2 with no spares and that's the safest way to get the maximum amount of disk on controller 1. Do you mean a RAID-4 aggregate on controller 2? I didn't think you can do a 2-disk RAID-DP aggregate.
|
|
#
?
Mar 18, 2013 23:07
|
|
|
Just do 9+3 and save everyone the trouble.
|
|
#
?
Mar 19, 2013 00:44
|
|
|
evil_bunnY posted:Just do 9+3 and save everyone the trouble.
|
|
#
?
Mar 19, 2013 02:31
|
|
|
madsushi posted:Do you mean a RAID-4 aggregate on controller 2? I didn't think you can do a 2-disk RAID-DP aggregate. Aye, neural misfire, I meant to type RAID-4. evil_bunnY posted:Just do 9+3 and save everyone the trouble. On systems this small sometimes every little bit counts and that extra data disk in a 10+2 config makes it more viable. If you don't need that bit of extra space or IO then 9+3 is certainly the easiest config to roll with, and will allow you to set the minimum spares option to 0 so the filer won't complain about not having a spare on the "passive" controller. That said, I really don't like the 12 disk systems and wish that 24 disks was the minimum you could get with an HA-pair.
|
|
#
?
Mar 19, 2013 04:36
|
|
|
adorai posted:His HA partner needs to own at least 2 disks. When I ordered the last bunch of disks I got 13 so I could have 2 raid groups and a spare. Surely you can find some budget for 1 drat disk?
|
|
#
?
Mar 19, 2013 10:50
|
|
|
evil_bunnY posted:Yes 3 is more than 2. Not when the shelf only holds 12. I guess you mean having a spare available sitting in a box somewhere?
|
|
#
?
Mar 19, 2013 12:58
|
|
|
Crackbone posted:Not when the shelf only holds 12.
|
|
#
?
Mar 19, 2013 14:11
|
|
|
The 2240 would have come with 24 bays but the 2220 only has 12.
|
|
#
?
Mar 19, 2013 14:24
|
|
|
evil_bunnY posted:I didn't realize this was the case. My shelves hold 24, and there's 13 disks in the second one. It's the difference between the shelves (which all hold 24 now) and the head/internal shelf (which holds 12 in the 2220 and 24 in the 2240). The 2240-4 is actually a shelf with filer controllers in the back; you can swap them out for shelf controllers later during an upgrade.
|
|
#
?
Mar 19, 2013 16:49
|
|
|
Ye, our 2240 is all-24. I didn't know the 2220 came with 12.
|
|
#
?
Mar 19, 2013 17:22
|
|
|
Any recommendations on storage solutions to look at for cheap mass storage? We're a NetApp shop but even a 2240 loaded out with 3TB drives is beyond the price-point for a lot of the groups I work with (higher ed, lol). I'm looking for something that can do NFS/CIFS and won't drive me insane operating it for essentially as cheap as possible. Has anyone bit the bullet on clustered-ontap yet? I just put our first cluster into production last weekend and I wouldn't describe it as "smooth sailing".
|
|
#
?
Mar 28, 2013 18:26
|
|
|
Clustered ONTAP, at the moment, is probably a mistake to deploy in production. There are a lot of missing features (SnapVault, 7-mode to cluster-mode replication, etc) and there is no easy upgrade path for existing 7-mode environments. I am a pretty big NetApp fan but I won't be touching cluster-mode for at least another year or two, when there's a transition model that doesn't involve "buy all new gear". How many environments are really suffering because they can't scale-out? The right choice is usually scale-up (new head). If you're a large enough environment to really care about scale-out, you probably have an SLA that doesn't make moving to a relatively new OS/paradigm that tolerable. My concern is that everyone at NetApp (from sales to marketing) has been instructed to sell/advocate Cluster-Mode and so NetApp customers/prospects are getting blitzed with cluster-mode information and then our customers ask us about it and once we explain that they can't do their SnapVault backups and that it will require a rip-and-replace upgrade, they wonder why NetApp even brought it up.
|
|
#
?
Mar 28, 2013 18:52
|
|
|
parid posted:Has anyone bit the bullet on clustered-ontap yet? I just put our first cluster into production last weekend and I wouldn't describe it as "smooth sailing". Can you describe the environment where you deploying cluster mode to? Like madsushi said, it seems like Netapp keeps pushing cmode even in situations where it's obviously not needed.
|
|
#
?
Mar 28, 2013 20:06
|
|
|
parid posted:Has anyone bit the bullet on clustered-ontap yet? I just put our first cluster into production last weekend and I wouldn't describe it as "smooth sailing". We're in the process now, it seems like the reason the guy in charge of the project came up with to do it instead of just going 7-mode was "because". I don't have super warm/fuzzies, so far I'm not really seeing any actual benefits, we don't have more than 2 heads in any of our sites at the moment.
|
|
#
?
Mar 28, 2013 20:19
|
|
|
The biggest driver for me was the lack of development for 7-mode. When we buy controllers, we're making a 5 year investment. It was a choice between staying with 7-mode and risking never having cifs3/nfs improvements. I bet that it would be worth the growing pains now for the pay off later when its stable. The big challenge to all of this is that all we purchased was a controller uplift, no new disks. We were given loaner trays as "swing gear" to allow the migration to happen. I agree with the rest of you. Sales is pushing it too hard and too early. It's an awkward place for netapp to be in. Abandoning further development on their stable proven platform and betting on a system that isn't ready yet. I'm getting tired of hearing "such and such feature should be available in 8.2!". 8.2 looks to be the release that would normally be considered production ready. That would be my advice to anyone thinking about clustered ontap, wait for 8.2. ~70 volumes and 60tb of sata with flash cache cards on a pair of 3250s. This particular system is in a jack of all trades role currently just CIFS and NFS. No block level storage. Mostly hosting things for campus, some lower level VM hosting, home drives. Of all our controllers its the most likely to benefit from the big promises of clustered ontap (unified name space, shifting data around, etc). There are a lot of really neat things. The new CLI is awesome. The unified namespace is great. The concept of export-policies is very handy too. There is a replication tool for migration. Currently, you have to engage PS and get them to bring it. We did and it was worth it. It's volume based. It essentially queries the source filer, does a sanity check, checks on many existing cifs/nfs settings, and sets up a snapmirror relationship with a cluster node. When you go to finish the migration it breaks off the mirror, sets up the shares/export-policies and away you go. Of the 70 volumes I moved it with it, I only had problems with 3. It only took 15 minutes to clean them up by hand. This tool is likely the basis of what the fancier, user accessible tool will be (which haha I hear will be released around 8.2). We managed to move everything with about 3.5 hours of downtime.
|
|
#
?
Mar 28, 2013 23:28
|
|
|
parid posted:Any recommendations on storage solutions to look at for cheap mass storage? We're a NetApp shop but even a 2240 loaded out with 3TB drives is beyond the price-point for a lot of the groups I work with (higher ed, lol). I'm looking for something that can do NFS/CIFS and won't drive me insane operating it for essentially as cheap as possible.
|
|
#
?
Mar 29, 2013 00:30
|
|
|
How cheap? Back blaze 1PB storage? 2x Synology RS10613xs+ with base 40TB storage looks like a reasonable option, add shelves to expand as require although 424TB sounds a little crazy.
|
|
#
?
Mar 29, 2013 00:46
|
|
|
You could also get a ReadyNAS 4200. That's 24TBish raw. You'll also want to cut off your own balls with a rusty razor afterward but that's neither here nor there.
|
|
#
?
Mar 29, 2013 00:48
|
|
|
Something well under $1,000 / TB usable. I hate to not buy HA in a consolidated storage system, but I think its on the table to negotiate away if its cheap enough. This is the kind of storage that will attempt to be pitched to people who run external HDs full of research data and who think a student, an external HD, and sneaker net is a data protection strategy. Sure, it will never be that cheap but we have been unsuccessful trying to get them into enterprise storage in one bite. This would also be a place to keep archive data. It would probably need to grow into the 200TB range without being a management nightmare (like 14 15TB systems might be). Some good ideas already, I'll check them out. How is oracle to work with for storage? I have had nothing but horrible experiences with their other business units.
|
|
#
?
Mar 29, 2013 01:16
|
|
|
parid posted:Some good ideas already, I'll check them out. How is oracle to work with for storage? I have had nothing but horrible experiences with their other business units.
|
|
#
?
Mar 29, 2013 01:53
|
|
|
We're around $125/TB for our LTO tier on SONAS, and that includes a backup copy of the data and disk to handle 10% rate-of-change before it's spooled off to tape. Scaling up and not down is often a much smarter way to save money. Higher up-front cost, though. Vulture Culture fucked around with this message at 04:04 on Mar 29, 2013 |
|
#
?
Mar 29, 2013 03:59
|
|
|
Hey, Clustered OnTAP stuff, neat. I'm neck deep in this stuff right now because my customer is looking at cDOT for their Exchange environment somewhere in the next year and NetApp is pushing PSEs very hard to get certified and comfortable with it. I did a 4 node cluster install a few weeks ago for my demo environment. So here are some of my thoughts: - cDOT is a better OS than DOT. There was some cleanup done to get rid of cruft in the DOT code. The CLI is significantly improved. The multi-tenancy support is much more fleshed out than something like multistore. The complete separation of vservers, as a logical entity, from nodes and aggregates makes things a lot better as you no longer have two distinct filers with two distinct personalities that just happen to be in an HA pair together. I like it a lot more than DOT. - The actual cluster setup is very straightforward and easy. Plug some cables in, run the cluster setup wizard on all nodes, and you've got a cluster. The configuration of vservers, network interfaces, protocols, namespaces and the rest is much more complicated and won't necessarily be trivial to 7-mode admins. Professional Services assistance is probably required, at least early on during migrations. In particular understanding how position in the namespace interacts with export policies can be confusing. The documentation that discusses user-mapping and CIFS setup is also pretty bad and makes it seem much more complicated than it really is to set up a CIFS share. - There is a lot of feature parity between 7-mode and cDOT already and 8.2 will close the gap even more. 8.2RC1 should drop in April with a GA shortly after. 8.2 includes snapvault like functionality (but better because it will take advantage of space efficiencies like compression and dedupe). A customer based 7-mode to cDOT migration tool should also be coming in the same timeframe as 8.2, so professional services will not longer be required. There should also be more flexibility in migrating 7-mode snapmirror relationships over to cDOT systems. And 2-node switchless clusters. Lots of good stuff in 8.2 which I think will make it more palatable for customers that have been holding off. - Cluster failover times are much improved on cDOT. This is big for some customers that have a lot of volumes or very busy controllers and can end up with long takeover/giveback times. Because nodes only have to take ownership of disks, and don't have to take over the whole personality of the HA partner there is much less work to be done during takeover/giveback. iSCSI is also better behaved during takeover since MPIO is used and network interfaces never fail over. More setup is required though, as servers should have a path configured to each node in the cluster. - To me, the major benefit of cDOT isn't scale out, it's the idea that you'll never have to take downtime again. It's always been easy to add compute power by buying larger arrays or more arrays to split workload between. cDOT doesn't fundamentally change that. But being able to move volumes non-disruptively between aggregates and add/remove nodes from the cluster on the fly you could upgrade your entire environment from 3210s to 6250s without ever taking an outage. You can also add disk anywhere and move workload to it without the customer ever being the wiser. If you want to retire an old disk shelf technology or you've just run out of space in an aggregate and can't grow it any more you don't hate to take outages to move that data. Likewise if you have a saturated network port hosting multiple interfaces you can easily turn up another link and move one or more of those interfaces onto the new port non-disruptively. Lots of very cool stuff can be done. - If you're holding out for an in place upgrade option then you may have a long wait. PM me about this if you have questions. - 7-mode is essentially in maintenance only mode at this point. cDOT will be getting new features like QOS, further enhancements to the replication engine, and further development on tools like VSC, OnCommand, etc. 7-mode development is basically limited at this point to fixing bugs. That's not a compelling reason to upgrade in and of itself, but if you're already doing a hardware refresh and you want to be future proofed you should consider cDOT. I think cDOT is pretty neat. It's unfortunate that there isn't a cleaner migration path, but I do understand the push to get customers on cDOT. NetApp can't afford to spend the next 10 years fully supporting both 7-mode and cDOT. They need to get the majority of their customer base on only one OnTAP version as quickly as possible so that engineering resources aren't divided, and obviously cDOT is going to be that release.
|
|
#
?
Mar 29, 2013 04:27
|
|
|
The "standard" documentation is available and is mostly complete but the "community" documentation like forums, KBs, and other experiential knowledge shared from other users is pretty thin. All the bugs I have hit, and KBs to work around them are hidden from the public as well. NippleFloss, you seem pretty close to this, do you have any recommendations on where to get some of the community information? I have found googling incredibly difficult as cDOT/c-mode/clustered ontap/cluster mode has so many names and there is so much out there on 7-mode already. Also, fun thing you can do in cDOT is mount a volume from a completely different HA pair inside the file structure of another volume. That Archive directory in your home folder? That actually points to a different filer with cheaper disk. I'm not sure how I would use it yet, but its neat  .
|
|
#
?
Mar 29, 2013 16:36
|
|
|
I am always excited to read a NippleFloss post on this thread, love talking shop with smart guys.NippleFloss posted:There is a lot of feature parity between 7-mode and cDOT already and 8.2 will close the gap even more. 8.2RC1 should drop in April with a GA shortly after. 8.2 includes snapvault like functionality (but better because it will take advantage of space efficiencies like compression and dedupe). A customer based 7-mode to cDOT migration tool should also be coming in the same timeframe as 8.2, so professional services will not longer be required. There should also be more flexibility in migrating 7-mode snapmirror relationships over to cDOT systems. And 2-node switchless clusters. Lots of good stuff in 8.2 which I think will make it more palatable for customers that have been holding off. All of my NetApp customers are using SnapVault for backups. Literally every single client. It's one of the biggest value-adds: NetApp handles all the backups for SQL/Exchange/VMware and handles all of the replication. There's really no way for them to ever move to cDOT (is that the preferred acronym?) without SnapVault. And SnapVault still isn't available in cDOT. Yes, it's coming soon. Yes, the feature parity is definitely better now than it was a year ago when cDOT only supported like two protocols. But why aggressively push customers toward a product that doesn't meet their needs? Most of my customers have autosupport on, I am sure NetApp can see that they're using SnapVault. I called my customers to tell them about ONTAP 8.0 and all of the amazing features (16TB dedupe limit, DataMotion for volumes) when it was ready to deploy on their systems. NetApp has been calling them for the past year when the solution doesn't make sense. Example: One client has a 2040 in production and a 2020 at the DR site. NetApp just told them to buy a new filer to replace the 2020 because cDOT was worth it. Even though the 2020 (or its replacement) would be pointless without SnapVault. Not blaming this on you (of course) but the fact is that "feature parity... soon" should not equate to "make all of our customers move right now at any cost (to them)". NippleFloss posted:Cluster failover times are much improved on cDOT. This is big for some customers that have a lot of volumes or very busy controllers and can end up with long takeover/giveback times. Because nodes only have to take ownership of disks, and don't have to take over the whole personality of the HA partner there is much less work to be done during takeover/giveback. iSCSI is also better behaved during takeover since MPIO is used and network interfaces never fail over. More setup is required though, as servers should have a path configured to each node in the cluster. I definitely understand this story, and it's part of what makes cDOT attractive in the first place. What NetApp's really been good at is allowing you to abstract away from the disk level so that you only care about the presentation level. DataMotion for Volumes is a great example. NippleFloss posted:If you're holding out for an in place upgrade option then you may have a long wait. PM me about this if you have questions. In-place upgrades are another area where it feels like NetApp gets it wrong. I got stuck with a lot of 32/64-bit mismatched volumes and because you can't force an upgrade from a 32-bit to a 64-bit aggregate without adding disk (why?), I ended up not being able to SnapMirror those volumes anymore and had to move to SnapVault. NippleFloss posted:7-mode is essentially in maintenance only mode at this point. cDOT will be getting new features like QOS, further enhancements to the replication engine, and further development on tools like VSC, OnCommand, etc. 7-mode development is basically limited at this point to fixing bugs. That's not a compelling reason to upgrade in and of itself, but if you're already doing a hardware refresh and you want to be future proofed you should consider cDOT. I am real glad that they are still selling 7-mode on their filers then. It's sort of like switching horses midstream, except that your new horse is like 10 feet behind your current horse. Then you tell your old horse to stand in the stream and die. Or something. NippleFloss posted:I think cDOT is pretty neat. It's unfortunate that there isn't a cleaner migration path, but I do understand the push to get customers on cDOT. NetApp can't afford to spend the next 10 years fully supporting both 7-mode and cDOT. They need to get the majority of their customer base on only one OnTAP version as quickly as possible so that engineering resources aren't divided, and obviously cDOT is going to be that release. Everyone understands the push: it makes it easier for NetApp. But what I don't get is how it's acceptable for it to be so bad for the customer. Once you're at NetApp's size, you have to start worrying about your existing customers and not just net-new. How are any of my customers going to get to cDOT easily? Right now, they can't move at all. After 8.2, they might be able to, but probably only with PS bringing in enough temporary hardware for them to do the storage equivalent of a 3-point turn. After that, someday (maybe????) they can just upgrade what they already have? When the only two migration paths are: buy all new gear (or) hope that your rep gives you a good deal on a lot of PS work, why move?
|
|
#
?
Mar 29, 2013 16:54
|
|
|
parid posted:The "standard" documentation is available and is mostly complete but the "community" documentation like forums, KBs, and other experiential knowledge shared from other users is pretty thin. All the bugs I have hit, and KBs to work around them are hidden from the public as well. NippleFloss, you seem pretty close to this, do you have any recommendations on where to get some of the community information? I have found googling incredibly difficult as cDOT/c-mode/clustered ontap/cluster mode has so many names and there is so much out there on 7-mode already. Most of the community knowledge you're looking for is tribal knowledge within NetApp. The customer base is still small, relative to the 7-mode base, and all installations and deployments are driven by PS right now so there isn't a lot of customer tinkering going on. It's going to be a while before there are enough customers doing enough hands on work with cDOT to build up anything like what exists for 7-mode. The good news is that just like in 7-mode there are a lot of tools in cDOT that make it possible to view the low level functioning so it will get there. It's a chicken and egg problem though. NetApp needs more customers to move to cDOT so that a knowledge base can be built up, but customers don't want to move to it until that knowledge base is there because they have been spoiled by the wealth of information available for 7-mode. In the meantime your best recourse is just to open a ticket for anything and everything you come across. If you need to know a best practice, open a ticket. If you hit a bug, open a ticket. If you want NetApp to buy you lunch, open a ticket. Tickets for new issues often end up turning into burt fixes or documentation requests that eventually make their way to the support site and become publicly viewable. You can also send me a PM if you have any questions. I'm not nearly as good with cDOT as I'd like to be, but I've got a lab environment that I can test some things in and I can ping internal groups with questions I don't have an answer to. madsushi posted:Not blaming this on you (of course) but the fact is that "feature parity... soon" should not equate to "make all of our customers move right now at any cost (to them)". I agree. I've had to yank the reins on the account team I work with to get them to slow down on the rush to cDOT (there was an internal memo a while back that laid out the different levels of acceptable abbreviations and when to use each and where in stunningly silly detail. Marketing run amok. But yes, cluster mode is out of favor and "clustered Data OnTAP, "clustered OnTAP" and "cDOT" are the preferred usages, in that order of preference). I think the real push has come with 8.1 and the inclusion of SAN functionality as that has made it "good enough" feature-wise for many customers. It's still early and my *HOPE* is that the idea is to start the conversation early so that many customers are comfortable with it a year or so down the road when 8.2 has a few minor revisions under it's belt and feature parity is near 100%. That has been my goal with my customer. Start talking to them about it early on and get them comfortable with the idea, get them asking questions and thinking it through, so that they are ready for cDOT when it is ready for them. I think a lot of guys on the sales side are over-zealous though, or are being pushed to be over-zealous by management who badly wants to get more cDOT equipment out there so they can point to it and say "lots of other people are doing it, why aren't you?" But really, if SnapVault is your big sticking point then 8.2 deserves a fresh look once it goes GA. The replication engine for cDOT is worlds better than the NDMP based SnapVault engine on 7-mode. My customer uses VSM exclusively because the inability of SnapVault or QSM to handle space efficient data (de-duped or compressed) without re-inflating it doubled or tripled our replication times for some large data sets and broke our SLAs. Being able to use SnapVault like functionality to maintain different numbers of snapshots on the source and destination, and between different OnTAP versions, while also taking advantage of space efficiencies will be very nice. quote:In-place upgrades are another area where it feels like NetApp gets it wrong. I got stuck with a lot of 32/64-bit mismatched volumes and because you can't force an upgrade from a 32-bit to a 64-bit aggregate without adding disk (why?), I ended up not being able to SnapMirror those volumes anymore and had to move to SnapVault. In place aggregate upgrades came at the same time as heterogenous snap mirror so even if you had mismatched volumes you should still have been able to mirror between 32 and 64 bit volumes starting in 8.1. There are a few use cases where you might want to upgrade a 32-bit volume to 64-bit without adding disk (no Protection Manager support for heterogenous snapmirror for months after 8.1 was a big one for me) but in general that was much more painful than this will be. This is more like the move from tradvols to flexvols. No in place upgrade of any kind (though one was promised there and never delivered) and you'll just have to move your data. It created a lot of unhappy customers but eventually it happened and tradvols are now a thing of the past. The cDOT moves are probably going to be the same way. It unfortunate, but I also can't think of a single storage vendor other than NetApp that has used the same monolithic OS for the past 30 years without ever requiring their customers to truly MOVE their data to new hardware. Basically since it's inception you have been able to do in-place upgrades and head-swaps to get new features. That's a drat good track record but it couldn't last forever. As long as there are very good customer usable migration tools (which there aren't right now) that meet the needs of a majority of customers, and as long as sales teams are able to bundle in things like loaner shelves to swing data over like parid got then I think it will be less painful than imagined. Not ideal, but acceptable. quote:I am real glad that they are still selling 7-mode on their filers then. It's sort of like switching horses midstream, except that your new horse is like 10 feet behind your current horse. Then you tell your old horse to stand in the stream and die. Or something. Make no mistake, cDOT is really a much better storage platform so it's not really fair to say that it's "behind" 7-mode. It just doesn't have all of the features of 7-mode, but it also has a lot of stuff, even now, that 7pmode doesn't. It's more like you're changing from a horse to a car midstream and right now all you have is hay, and no gas, and you don't really know how to drive the car either and sometimes the car dies on you for no good reason (probably because you're driving it through a stream) whereas the horse has always gotten you where you needed to go and you get along well with it and understand it's many idiosyncrasies. Eventually you'll end up with the car, but and a while after that you'll never understand how you lived with the horse for as long as you did, but that's all going to take a while. quote:Everyone understands the push: it makes it easier for NetApp. But what I don't get is how it's acceptable for it to be so bad for the customer. Once you're at NetApp's size, you have to start worrying about your existing customers and not just net-new. I really couldn't tell you what the strategy is at the upper management level but I know from experience that currently there are not enough engineering resources to effectively continue development on both 7-mode and cDOT. I've had plenty of issues over the last few years go unfixed or wait a long time for a fix because the lions share of engineering resources have been devoted to getting cDOT ready for prime time. The longer that goes on the more unhappy customers there will be, so it seems like a case of short term pain for long term gain. Whether that calculation is correct remains to be seen, but the company is in a tricky spot right now where there is really no right answer short of "painless in-place upgrades" which doesn't seem feasible for technical reasons. quote:How are any of my customers going to get to cDOT easily? Right now, they can't move at all. After 8.2, they might be able to, but probably only with PS bringing in enough temporary hardware for them to do the storage equivalent of a 3-point turn. After that, someday (maybe????) they can just upgrade what they already have? When the only two migration paths are: buy all new gear (or) hope that your rep gives you a good deal on a lot of PS work, why move? They'll probably get to cDOT the same way mine will, by forcing NetApp to sweeten the deal on a hardware refresh to the point that it makes more sense to move than to not move. I expect that for customers looking to transition NetApp will be willing to eat the cost on some additional storage or PS hours if it means converting an account over to cDOT. That's really the only way it's going to work in a lot of accounts, so they'll either do that or fail miserably and go out of business. For the sake of my paycheck I hope it isn't the latter. Don't think I'm disagreeing with you here, because I'm not. The big push to move cDOT equipment for existing customers should have waited until 8.2 at the earliest and people on the sales side need to have a much better understanding of customer needs before they start trying to sell cDOT equipment. But I understand what the company is trying to do, and why it needs to do it.
|
|
#
?
Mar 29, 2013 18:51
|
|
|
NippleFloss posted:- If you're holding out for an in place upgrade option then you may have a long wait. PM me about this if you have questions. Ha, after the traditional -> flexvol migration, I just generally assume that there's never going to be a reasonable migration path, and do-over is the way to go.
|
|
#
?
Mar 29, 2013 19:34
|
|
|
Hey other guy that owns a starboard check this out. We are of course nervous with the systems we already have.  http://www.crn.com/news/storage/240150957/startup-starboard-storage-cuts-sales-staff-hangs-for-sale-sign.htm At least my boss is seriously talking to NetApp about their e-series products which would be a perfect fit for our needs.
|
|
#
?
Mar 30, 2013 03:25
|
|
|
We had a Snap Server 410 sitting in storage so my boss asked me to upgrade the storage and put it into use. I checked the official website: "Supports up to 2TB drives, hot swappable!" OK, so I filled it with 2TB drives, and used the old 500GB drives that were in it to build a VM server. Then I try to actually set this thing up. It turns out that unlike every other NAS I've encountered, this thing keeps its OS on the drives. The drives I just wiped. No biggie, I'll copy an image from the company's site. "Sorry," says phone support, "We don't have a process for that. The software comes preloaded on the drive in manufacturing. The only thing you can do is buy a whole new server." Are these loving guys for real? Looks like the best thing I can do is call around to our other facilities and see if someone else has a hard drive with the OS so I can clone it.
|
|
#
?
Apr 2, 2013 23:06
|
|
|
Snap Servers are only good for storing media at the houses of engineers who decommission them.
|
|
#
?
Apr 3, 2013 16:26
|
|
|
OK, well if I have these four 2TB SATA drives, what would be a cost effective NAS to pop them into if I just need them to hold people's Homefolders (preseving NTFS permissions, AD integration).
|
|
#
?
Apr 3, 2013 19:32
|
|
|
Zero VGS posted:We had a Snap Server 410 sitting in storage so my boss asked me to upgrade the storage and put it into use. I checked the official website: "Supports up to 2TB drives, hot swappable!" If all else fails, I have a 4500 and can try to image the non share data. You'd likely have to put the drive in another linux box and set up the partitions and stuff. It is running an /old/ version of GuardianOS, so might not be ideal.
|
|
#
?
Apr 3, 2013 23:43
|
|
|
EoRaptor posted:Yeah, I was a bit off: I know I'm bringing this comment back from the dead, but does the latest EqualLogic firmware still require customer support assistance to remove a member from a group? I remember reading this thread and seeing it posted, and I'm 99% sure I remember seeing it in the release notes back when I upgraded our arrays to 5.2.3. Now that I'm on 6.something.something, I can't find any references to it, and I can't find anything online that states it either. The reason I'm asking is because I need to remove a member from our DR group, but I'm sure as hell not clicking that button until I'm 100% sure I know what will happen.
|
|
#
?
Apr 4, 2013 00:53
|
|
|
Boogeyman posted:I know I'm bringing this comment back from the dead, but does the latest EqualLogic firmware still require customer support assistance to remove a member from a group? I remember reading this thread and seeing it posted, and I'm 99% sure I remember seeing it in the release notes back when I upgraded our arrays to 5.2.3. Now that I'm on 6.something.something, I can't find any references to it, and I can't find anything online that states it either. I don't think that was ever the case? It was simply placing the member into another storage pool thus vacating data off of it, waiting for that to complete (hours to days), and then deleting the member. (I don't use Equallogic any longer, but that was the case in 5.x)
|
|
#
?
Apr 4, 2013 00:58
|
|
|
Zero VGS posted:OK, well if I have these four 2TB SATA drives, what would be a cost effective NAS to pop them into if I just need them to hold people's Homefolders (preseving NTFS permissions, AD integration). Buy yourself a motherboard/ram/cpu/psu/usb key/case from newegg for less than $500, install Openfiler 2.99 on the USB stick and set up the four drives in software RAID-5. Openfiler can be easily AD-integrated and mapped to within AD profiles or batch scripts.
|
|
#
?
Apr 9, 2013 05:47
|
|
|
Agrikk posted:Buy yourself a motherboard/ram/cpu/psu/usb key/case from newegg for less than $500, install Openfiler 2.99 on the USB stick and set up the four drives in software RAID-5. Is it possible to setup NTFS permissions on an NFS share though? I get that you can set unix permissions, and use UID Mapping / NIS but that's still not the same
|
|
#
?
Apr 9, 2013 08:16
|
|
|
theperminator posted:Is it possible to setup NTFS permissions on an NFS share though? I get that you can set unix permissions, and use UID Mapping / NIS but that's still not the same The Openfiler CIFS implementation is pretty solid and allows you to directly map AD objects. If you have an old PC lying around, I suggest installing OF and giving it a whirl. The install takes about ten minutes and getting it AD integrated takes another 15 once you get the hang of it. There is a bit of a learning curve in navigating the management web site to figure out how they laid everything out, but check it out.
|
|
#
?
Apr 9, 2013 15:53
|
|
|
Awesome, this could come in quite handy. thanks!
|
|
#
?
Apr 10, 2013 08:16
|
|
|
|
| # ? Apr 26, 2024 04:26 |
|
|
We just got an Isislon cluster and I was going to be there while the tech would set it up in the rack (this is at a remote site across the country). They shipped the wrong Infiniband cables so not only did they overnight more cables they also are paying for me staying out longer than I needed to. The new cables were the same ones and were wrong again. Is this typical?
|
|
#
?
Apr 11, 2013 04:01
|
|
