|
Wicaeed posted:Ah, I figured it out. I had to disconnect all the current sessions and reconnect, then rescan for disks to see the newly created LUN drat Microsoft, drat!
|
#
?
Jan 26, 2012 19:56
|
|
|
|
| # ? Apr 27, 2024 14:56 |
|
|
Misogynist posted:The bigger losses by far are quotas
|
|
#
?
Jan 26, 2012 20:18
|
|
|
Wicaeed posted:Ah, I figured it out. I had to disconnect all the current sessions and reconnect, then rescan for disks to see the newly created LUN You should definitely not have to disconncet the current sessions to see a new LUN. That's not a feature limitation of the software iscsi initiator for sure.
|
|
#
?
Jan 26, 2012 20:34
|
|
|
Wicaeed posted:Ah, I figured it out. I had to disconnect all the current sessions and reconnect, then rescan for disks to see the newly created LUN
|
|
#
?
Jan 26, 2012 21:08
|
|
|
Hah, I found some awesome relic on YT: http://www.youtube.com/watch?v=RPqKvp0OWoM&feature=player_embedded!
|
|
#
?
Jan 26, 2012 21:26
|
|
|
szlevi posted:Hah, I found some awesome relic on YT: http://www.youtube.com/watch?v=RPqKvp0OWoM&feature=player_embedded! Is it bad that I want to punch that kid in the face?
|
|
#
?
Jan 26, 2012 21:32
|
|
|
FISHMANPET posted:Is it bad that I want to punch that kid in the face?  TBH I was more annoyed by his utter junk mouse/touch/ball/whatever pointer device that clicks louder than a hard drive after servo failure.
|
|
#
?
Jan 26, 2012 21:44
|
|
|
szlevi posted:
Yeah, that was part of it. E: The worst part for me is that nobody should be running that except developers and other Windows experts (not in a legal sense, in a what possible use could you have for that sense). Except he's just another lovely ricer that disables his page file or whatever other stupid poo poo people to do make their computers "faster."
|
|
#
?
Jan 26, 2012 21:53
|
|
|
marketingman posted:So let me stop you right there because there in your scenario there is no performance hit from dedupe or thin provisioning on a NetApp filer. This is from a few pages ago, but there is a performance hit with dedupe. You can't run realloc on a deduped volume, so after a while you start to get an uneven distribution of data across your disks, which leads to a loss in performance. Results may vary. By the by, for anyone running Netapp with less than 8.1 firmware, you should run the realloc command on each volume every time you add disks to your aggregate.
|
|
#
?
Jan 28, 2012 18:18
|
|
|
What's preventing you from turning it off to realloc in your maintenance window?
|
|
#
?
Jan 28, 2012 18:39
|
|
|
Nomex posted:This is from a few pages ago, but there is a performance hit with dedupe. You can't run realloc on a deduped volume, so after a while you start to get an uneven distribution of data across your disks, which leads to a loss in performance. Results may vary. Uhh... You can run reallocate on a deduped volume, and you can run reallocate on a snapshotted volume, you just have to throw the right flags to the reallocate command. Dedupe has two areas where it can lower performance: during your nightly dedupe window (which can be seconds to minutes based on your daily rate of change) and new writes can take slightly longer. Dedupe can also improve performance, by letting you get more unique blocks in your cache.
|
|
#
?
Jan 28, 2012 19:00
|
|
|
madsushi posted:Uhh... There is a third area where ASIS dedupe *CAN* lower performance, but it's harder to quantify. In a heavy sequential read environment dedupe will artificially fragment otherwise sequential groups of blocks if those blocks are shared across files. In this case a sequential read of a file turns into a random read on disk and will perform slower. The caveat is that dedupe aware caching will often come into play in those same workloads, so the two effects can cancel each other to a degree. I still generally recommend avoiding enabling dedupe on heavily sequential workloads where performance is important. Database transaction logs would be one obvious one as slow trans log performance can affect the speed of database interactivity, and generally the benefit of deduping transaction logs is small. Another workload would be something like streaming video or audio where you will likely be reading back large, contiguous segments of files. For most workloads, though, dedupe has no noticable performance impact and the space savings can be significant.
|
|
#
?
Jan 28, 2012 21:10
|
|
|
Curious how different implementations of the 'cp' command works when copying from one network location to another. Do the bits actually have to go through the machine running the command? Are there any common or budget implementations where they do not?
|
|
#
?
Jan 29, 2012 01:30
|
|
|
Shaocaholica posted:Curious how different implementations of the 'cp' command works when copying from one network location to another. Do the bits actually have to go through the machine running the command? Are there any common or budget implementations where they do not? Yes. If the devices support NDMP, you can run something like NDMPCopy to have it do a system to system transfer.
|
|
#
?
Jan 29, 2012 02:12
|
|
|
Not sure if this is the right place to ask this but I am getting desperate... I am currently using Oneiric and I am trying to set up juju with my ec2 instance with s3 storage to run rtgui and torrent to the cloud. I am having trouble running "juju bootstrap" in terminal. It says I do not have the right public key. I believe the parameter in "environments.yaml" this is supposed to be represented in is "admin-secret:" How do I find this? I cannot find anything about a public key in my aws console... Anyone done anything similar? Tutorials? I am trying to follow this tutorial but I'm getting lost. http://cloud.ubuntu.com/2011/09/torrent-download-cloud-appliance/
|
|
#
?
Jan 29, 2012 10:22
|
|
|
I'd also point out I specifically said no performance hit in that environment. Two VMware essentials servers? Do we really think we're going to stress even a low end FAS with that? Nipple Floss is correct on the sequential workload and TBH it's something I forgot due to PAM, as he also mentioned. Edit: Also can I get a show of hands of people regularly running wafl reallocate commands? This seems highly unusual to me.
|
|
#
?
Jan 29, 2012 13:17
|
|
|
madkapitolist posted:Not sure if this is the right place to ask this but I am getting desperate... You didn't even bother to look for a solution, did you? Juju Tutorial, first page posted:Note If you already have an AWS account, you can determine your access key by visiting http://aws.amazon.com/account, clicking “Security Credentials” and then clicking “Access Credentials”. You’ll be taken to a table that lists your access keys and has a “show” link for each access key that will reveal the associated secret key. From here: https://juju.ubuntu.com/docs/getting-started.html
|
|
#
?
Jan 29, 2012 13:56
|
|
|
quote:
Cairo and WinFS were supposed to be object stores. Not sure what the plan for Cairo was, being that old, but WinFS wasn't even a filesystem. It was a masked SQL Server instance with some ORM, running a bunch of database files hidden away in C:\System Volume Information. Two of the biggest reasons it failed was because they couldn't get FileStreams out on time in SQL Server back then, and that its WinFS' main APIs were pretty much only .NET. If something like WinRT would have existed back then, it'd have made more sense, seeing how .NET established itself for consumer application development (note: not really). Anyway, ReFS is just a block filesystem like NTFS. Same upper level APIs, different on-disk format. It's completely B+ tree now and can do copy-on-write. As far as the other cut features go, I figure those that still make sense will slowly be added back. One of the reasons why it won't ship to the client yet.
|
|
#
?
Jan 29, 2012 14:08
|
|
Cairo, WinFS... ReFS?
|
ZeitGeits posted:You didn't even bother to look for a solution, did you? I've gone through that before. from the .yaml file environments: sample type: ec2 access-key: from security credentials page secret-key: from security credentials page control-bucket: from security credentials page default-series: oneiric admin-secret: this is not "access" or "secret" key. The code generated by the X.509 certificate doesn't work either. Which one SHOULD it be?
|
|
#
?
Jan 29, 2012 17:56
|
|
|
madkapitolist posted:I've gone through that before. Sorry, I tend to assume the worst. No offense meant. Check this page: http://blog.taggesell.de/index.php?/archives/73-Managing-Amazon-EC2-SSH-login-and-protecting-your-instances.html Your problem sounds like a ssh key mismatch. Edit: I'm sorry for derailing this thread. Madkapitolist: Open a new thread in the tech support forum if your problem persists. ZeitGeits fucked around with this message at 18:32 on Jan 29, 2012 |
|
#
?
Jan 29, 2012 18:21
|
|
|
NippleFloss posted:There is a third area where ASIS dedupe *CAN* lower performance, but it's harder to quantify. In a heavy sequential read environment dedupe will artificially fragment otherwise sequential groups of blocks if those blocks are shared across files. In this case a sequential read of a file turns into a random read on disk and will perform slower. The caveat is that dedupe aware caching will often come into play in those same workloads, so the two effects can cancel each other to a degree. I'm not really sure if this argument makes sense on a NetApp though, due to WAFL, there's really no such thing as "sequential" since the data can be anywhere.
|
|
#
?
Jan 29, 2012 19:39
|
|
|
Combat Pretzel posted:Not to spoil your ReFS bickering, but Cairo and WinFS have poo poo to do with ReFS. You didn't spoil anything because you clearly didn't get the point. quote:Cairo and WinFS were supposed to be object stores. Not sure what the plan for Cairo was, being that old, but WinFS wasn't even a filesystem. It was a masked SQL Server instance with some ORM, running a bunch of database files hidden away in C:\System Volume Information. Two of the biggest reasons it failed was because they couldn't get FileStreams out on time in SQL Server back then, and that its WinFS' main APIs were pretty much only .NET. If something like WinRT would have existed back then, it'd have made more sense, seeing how .NET established itself for consumer application development (note: not really). What you're talking about here is the last iteration before it's got whacked - but, as I wrote, WinFS started out well before that, as a fs (hence its name, y'know.) quote:Anyway, ReFS is just a block filesystem like NTFS. Same upper level APIs, different on-disk format. It's completely B+ tree now and can do copy-on-write. As far as the other cut features go, I figure those that still make sense will slowly be added back. One of the reasons why it won't ship to the client yet. yeah, it's a rather crappy patched-up NTFS - but it's still the same idea except now they make even less effort to solve their inherent disadvantages when it comes to managing data; they just make some bulletpoint changes and that's it. It will go through at least 2 more generations before it gets usable and by then who knows where the world will be...
|
|
#
?
Jan 29, 2012 23:16
|
|
|
szlevi posted:You didn't spoil anything because you clearly didn't get the point. szlevi posted:What you're talking about here is the last iteration before it's got whacked - but, as I wrote, WinFS started out well before that, as a fs (hence its name, y'know.) As for Cairo OFS, there's no public information how its development versions were implemented at all. Might just have been a similar layer on top of traditional NTFS as well. --fake edit: If I am to believe Wikipedia, their object store attempts were almost all based on SQL Server since at least 1995. So yeah, NTFS all the way. szlevi posted:yeah, it's a rather crappy patched-up NTFS - but it's still the same idea except now they make even less effort to solve their inherent disadvantages when it comes to managing data; they just make some bulletpoint changes and that's it. So I'm not sure what you mean by managing data. In what way does NTFS have disadvantages? szlevi posted:It will go through at least 2 more generations before it gets usable and by then who knows where the world will be...
|
|
#
?
Jan 30, 2012 00:17
|
|
|
madsushi posted:I'm not really sure if this argument makes sense on a NetApp though, due to WAFL, there's really no such thing as "sequential" since the data can be anywhere. That's not really true. WAFL buffers blocks to be written in memory and writes them out in a batch to disk. For performance reasons it tries to find a contiguous segment of the aggregate large enough to hold all of the blocks in the buffer so it can treat the write as a streaming write. If the actual write activity was truly random (meaning write a block here in this free space, and a block here in this free space) then they would never complete the writes before the next buffer flush was due. The reason WAFL traditionally has good write performance is because it turns all write activity, either random or sequential, into sequential writes on disk. As long as that data is written once and then not overwritten at random and in pieces then the data remains mostly sequential on disk. WAFL fragmentation happens when you write the data out sequentially and then overwrite it randomly. But for the sort of workloads that we are talking about, where it is a sequential write followed by a sequential read, with no overwrite, then fragmentation is not a problem. Additionally, there are commands like LUN reallocate, volume reallocate, read reallocate and extents. The goal of all of those commands are to relayout the LUN or volume to improve sequential read performance by grouping file blocks together.
|
|
#
?
Jan 30, 2012 00:36
|
|
|
Awesome, thanks for the cited history refresher! I watched some of that unfold as it happened, just stopped thinking about it so long ago... Also, on the bickering between Cairo/WinFS/ReFS, etc: I asked if this new ReFS that was linked was the one that was supposed to come out ages ago, (Cairo and/or WinFS) since I haven't been following it. szlevi gave a very nice list of articles informing me of the actual history of what I was asking about, so thanks. No need to run in the special olympics over it.szlevi posted:Yeah, Apple by default treats its customers like poo poo so they really didn't care what they think. At all. Intraveinous fucked around with this message at 18:18 on Jan 30, 2012 |
|
#
?
Jan 30, 2012 18:10
|
|
|
Martytoof posted:Well the fact that OpenSolaris is no more is kind of a wrench in things. Not to say they can't keep this poo poo working well with what they have, but it sort of limits where they can go with it. I dunno, from watching this presentation it sounds like Solaris was really lucky it got open sourced before the Oracle takeover, since most of the lead programmers on the cool solaris technologies quit rapidly in protest. Most of them are still working on their original projects, just getting paid by different companies. But yeah I don't think I'd be very comfortable running a production system on Nexenta, OpenIndiana or SmartOS just yet 
|
|
#
?
Feb 2, 2012 18:03
|
|
|
What's the deal with Equallogic? I seem to remember looking into their kit a couple years ago and there was some issue with the controllers not being active/active, or something to do with how they failed over. What are the EQL gotchas I should be looking at? They're probably going to push us towards Compellent, which is fine by me, but I don't sign the checks so that may not happen. Had another VAR pushing NexSAN quite heavily. Anyone have experience with them?
|
|
#
?
Feb 2, 2012 19:01
|
|
|
error1 posted:I dunno, from watching this presentation it sounds like Solaris was really lucky it got open sourced before the Oracle takeover, since most of the lead programmers on the cool solaris technologies quit rapidly in protest. This is pretty fascinating, thanks! Dude is hella spergy though. Who would have imagined!
|
|
#
?
Feb 2, 2012 19:27
|
|
|
error1 posted:But yeah I don't think I'd be very comfortable running a production system on Nexenta, OpenIndiana or SmartOS just yet I had my first scare with Nexenta in production the other day. I was trying to delete some crap in the Datastore Explorer and it stopped responding. VMs didn't want to shutdown and when they finally did, they wouldn't come back up. Restarting the host never quite finished. Fortunately the VMs that were on already still worked so I just waited until the maintenance window to restart the Nexenta box. I think it had something to do with thin provisioning and iSCSI, but it's working now. Definitely planning on getting real storage in the next budget.
|
|
#
?
Feb 2, 2012 19:52
|
|
|
Do any of you do SAN snapshots and replication of VMFS? Ever encountered any issues when you mount up a snapshot at a remote site?
|
|
#
?
Feb 2, 2012 20:26
|
|
|
ozmunkeh posted:What's the deal with Equallogic? I seem to remember looking into their kit a couple years ago and there was some issue with the controllers not being active/active, or something to do with how they failed over. What are the EQL gotchas I should be looking at? They're probably going to push us towards Compellent, which is fine by me, but I don't sign the checks so that may not happen. On the Equallogic side, the controllers are active/passive but as far as I know failover is not an issue. I can't ever remember having a problem in ~5 years, and if I did have a controller fail I didn't notice an impact. Bitch Stewie posted:Do any of you do SAN snapshots and replication of VMFS? Ever encountered any issues when you mount up a snapshot at a remote site? I do and no, I never have. Although I will say I do not do it that often, and I would be wary of relying on it for database applications (SQL, Exchange, etc). Internet Explorer fucked around with this message at 20:54 on Feb 2, 2012 |
|
#
?
Feb 2, 2012 20:52
|
|
|
Combat Pretzel posted:What point? You were equating it to the efforts of Cairo and WinFS. Which makes no sense, since it's not remotely comparable. ReFS isn't trying to be anything else than a block filesystem. OK, let me reiterate again: the purpose of all these gymnastics was and still is to provide better management of the information. The whole point in OFS (Cairo) was to introduce an object-based system and while WinFS was not a replacement in any way the aim of its development was focusing on the same thing (ie information management.) quote:--fake edit: If I am to believe Wikipedia, their object store attempts were almost all based on SQL Server since at least 1995. So yeah, NTFS all the way. Since you seem to be big on Wikipedia (I don't really trust it any more than a blog entry) here's your relevant part: http://en.wikipedia.org/wiki/WinFS#Development [/i]"The development of WinFS is an extension to a feature which was initially planned in the early 1990s. Dubbed Object File System, it was supposed to be included as part of Cairo. OFS was supposed to have powerful data aggregation features,[12] but the Cairo project was shelved, and with it OFS. However, later during the development of COM, a storage system, called Storage+, based on then-upcoming SQL Server 8.0, was planned, which was slated to offer similar aggregation features.[12] This, too, never materialized, and a similar technology, Relational File System (RFS), was conceived to be launched with SQL Server 2000.[12]However, SQL Server 2000 ended up being a minor upgrade to SQL Server 7.0 and RFS was not implemented. But the concept was not scrapped.[12] It just morphed into WinFS."[/i] Happy now? quote:Unless I'm missing something, every filesystem actually in use works practically the same. Hierarchically organized unstructured streams. If other semantics were of advantage over that, there would have been some big solutions established by now, if only in open source space. Let me guess, you have not seen any object-based storage system... but if you have seen then please explain how they would work running NTFS and why that's not possible... :P quote:Maybe I'm not seeing it, but latest developments in filesystems, like ZFS and BTRFS, don't attempt to be different. You mean except the fact that NTFS lacks almost all the features of the other ones? quote:So I'm not sure what you mean by managing data. In what way does NTFS have disadvantages? Are you serious? No volume management, no RAID, no checksums whatsoever etc etc? FYI unless you're some MS shill it should be common knowledge just how far NTFS is behind them but at any rate it's pretty simple to answer yourself just by using Google. quote:Probably the same place where it is now. Just with flashier UIs. If Microsoft doesn't pull another Longhorn, 2 generation's pretty much just 6 years. Since they've announced a three year cadence in Windows development. I am not sure if I understand your PoV - these announcements have nothing to do with UI, my comment was about the fact that MS really lags behind others and very slow to roll out new features and new versions.
|
|
#
?
Feb 2, 2012 21:13
|
|
|
Internet Explorer posted:I would be wary of relying on it for database applications (SQL, Exchange, etc).
|
|
#
?
Feb 2, 2012 22:03
|
|
|
evil_bunnY posted:Those two would have plugins to make the DBMS snap-aware though. Exchange does, we run SQL on VMDKs though so that doesn't. To put it in context, we do proper backups, but I'm just in the process of setting up a server that will hold replicas from our primary SAN that will go at the other side of our campus - essentially it's a "I'm in deep loving poo poo and need something to get the business out of the ashes before the box of tapes gets brought up from 150 miles away" box.
|
|
#
?
Feb 2, 2012 22:12
|
|
|
Q: I'm running a 2008 R2 cluster (Hyper-V), what's the best practice to create & run a handful of identical VMs?
|
|
#
?
Feb 2, 2012 22:13
|
|
|
evil_bunnY posted:Those two would have plugins to make the DBMS snap-aware though. True, forgot to mention that. Now a days you should be able to find plugins for VMware/XenServer/Hyper-V and all the major storage vendors to do proper snapshots. Otherwise you are just getting a crash consistent copy. Bitch Stewie posted:Exchange does, we run SQL on VMDKs though so that doesn't. There should be a plugin you can use that will give you a proper snapshot of SQL. I forget the term they use, but I think a coalesce snapshot? It basically puts the database in backup mode for a second, then takes the snapshot, usually using something like VSS.
|
|
#
?
Feb 2, 2012 22:16
|
|
|
Internet Explorer posted:There should be a plugin you can use that will give you a proper snapshot of SQL. I forget the term they use, but I think a coalesce snapshot? It basically puts the database in backup mode for a second, then takes the snapshot, usually using something like VSS. Quiesced? The SAN vendor makes one, but it only works if the DB/log volumes are hosted directly on the SAN, it won't work if they are VMDKs. Tbh, crash consistent VM backups would be just fine in the scenario this box is there to aid in, my main concern is how prone VMFS is to screwing up if you just snapshot it and then try and mount it.
|
|
#
?
Feb 2, 2012 22:23
|
|
|
Bitch Stewie posted:Quiesced? The SAN vendor makes one, but it only works if the DB/log volumes are hosted directly on the SAN, it won't work if they are VMDKs. Yes, sorry. Having a bit of a brain fart. I could have sworn there was a way to do it with a VMDK. I think VMFS is pretty resilient. I have used snapshots for test labs a lot and never ran into any problems, even Exchange or SQL.
|
|
#
?
Feb 2, 2012 22:31
|
|
|
Internet Explorer posted:Yes, sorry. Having a bit of a brain fart. I could have sworn there was a way to do it with a VMDK. I think VMFS is pretty resilient. I have used snapshots for test labs a lot and never ran into any problems, even Exchange or SQL. There are products that interact with VSS on the guest to get things into a nice happy state for snapshotting. To your point though, crash consistent snapshots are generally "good enough"
|
|
#
?
Feb 2, 2012 22:42
|
|
|
|
| # ? Apr 27, 2024 14:56 |
|
|
szlevi posted:Q: I'm running a 2008 R2 cluster (Hyper-V), what's the best practice to create & run a handful of identical VMs? Did you perhaps mean to ask your question in here?
|
|
#
?
Feb 2, 2012 23:25
|
|