Tag Info

Hot answers tagged

11

Much has happened since I asked this question in October 2010. As of September 2013, a new collaboration known as OpenZFS will serve as a central site for several ZFS projects. The new site is http://open-zfs.org/ (with a dash) Today at LinuxCon North America, Brian Behlendorf and Matthew Ahrens are announcing that members from the illumos, zfsonlinux.org, ...


9

A customer just asked me for this, so I compiled a short list of links I have found useful over the years. Not necessarily in the best order, and by no means complete, but thought it might make a good addition for anyone hitting this question: http://www.nex7.com/readme1st (work in progress, can PM me if you have questions) ...


8

I'll ask, how important is it that you specifically use KVM? My preference for the type of solution you're inquiring about is to build around VMWare ESXi. You can build an all-in-one server running VMWare ESXi booting from flash media (SDHC, USB, CF) and leverage the DirectPath I/O (PCI-passthrough) available on current servers to present a SAS/SATA HBA to ...


6

I've been using SmartOS and KVM in production for a few months now and am very happy with it. It sounds like it should suit your needs just fine. All the ZFS stuff you require is supported. For the monitoring stuff you'll need setup up some third party stuff though. I'm working a couple of projects related to monitoring and specifically for doing the things ...


6

It appears OpenIndiana does not use the shareiscsi property. Also (perhaps relatedly) COMSTAR is not included as part of the base installation. So I fired up a fresh OpenIndiana VM and ran the following commands and was able to create, share and mount a sparse LUN without any drama: sudo -s pkg install network/iscsi/target svcadm enable -s stmf svcadm ...


5

Oracle officially killed OpenSolaris, so the website will not be updated anymore. FreeBSD is working with the IllumOS project to advance ZFS at this point (their website is still a bit scant on details). I don't know if these changes are derived from actual Solaris 11 development or not, nor if the changes will be merged into the official project (a likely ...


5

Two things... The most important document for you to read is the Oracle Solaris ZFS Administration Guide. It provides a comprehensive description of ZFS features and terminology. It also gives excellent examples. Beyond that, you may want to look at NexentaStor as an option if you're primarily interested in the storage aspects of ZFS. It's a ...


4

The zfs mailing list remains on opensolaris.org, and Oracle has not established a new public website for their ongoing ZFS development project. Update (2013): The opensolaris.org website has now been shut down, and Oracle’s ZFS mailing list has moved to the new Solaris-ZFS.java.net website.


4

For versions up to 28, you can still browse the repository in OpenSolaris web site: http://src.opensolaris.org/source/history/onnv/onnv-gate/usr/src/uts/common/sys/fs/zfs.h http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/sys/fs/zfs.h#348 Update: This repository has moved to: ...


4

yes, via DTrace's NFS providers - http://wikis.sun.com/display/DTrace/nfsv3+Provider. Additionally you should be able to find various DTrace scripts around the net that offer NFS client statistics in a more detailed way than plain nfsstat.


4

The first script fails because the example is incorrect. You can't reference args[] while globbing the probe in that manner (as probes matching the pattern may come and go while the enabling is still in effect, so guarantees about the type of args[N] can't be made). You'd need to list each nfs3-op-* individually, comma separated ...


3

Your OpenIndiana system needs to have two network interfaces in order to accomplish what you're trying to do. In the vSphere client, add another Network adapter and address it on the 10.0.0.x subnet. Make sure that one network adapter is assigned to the "SAN" network label and the other to the "VM Network" label. One network adapter is so you can see and ...


3

If you're in a VT (Virtual Terminal), you can usually scroll with Ctrl+Page Up and Ctrl+Page Down. If you're in a GUI, there should be a scrollbar, or you can try the same key combination. In either environment, you can pipe the command to less, like so: zfs | less Within less, you can use Page Up, Up, Page Down, Down just as you'd expect. Q will ...


3

Well, I'm curious what your zpool setup looks like. As nice as L2ARC cache is, it does not seem like you have a robust pool design since you only have four disks. Another key is to determine if your workload could even benefit from L2ARC. How much RAM do you have in the system? The path of least resistance is to use an approved/tested SSD for L2ARC. I ...


3

by default add-view will create a view with the following properties: View Entry: 0 LUN : 0 Host group : All Target group : All If you want to define which targets map to a view/LU you need to create a target-group: stmfadm create-tg TG1 then add your target to it: stmfadm add-tg-member -g TG1 ...


3

You can remove all non trivial ACLs in ZFS with the following: chmod A- filename Source: Solaris ZFS Administration Guide: Using ACLs and Attributes to Protect ZFS Files


3

How about this... In this instance I have a ZFS filesystem '/volumes/lab3_pool_a/data02/alpha/' and I am using fsstat to get some basic statistics about it. Notice, I am getting stats for both, ZFS as a whole, and this one filesystem. By the way, fsstat has some good options and may give you just what you need. # fsstat zfs ...


3

I can tell you the CLI commands for creating the pool. I don't know the GUI or WebUI though. #replace drive0... with your drives zpool create tank drive0 drive1 #This will create a striped set, no redundancy, single failure = dead pool zpool create tank mirror drive0 drive1 mirror drive2 drive3 #This will create a stripe set of mirrors, basically ...


3

I don't have a Solaris system to hand but it appears that the script is relying on /bin/sh being a link to /bin/bash and as such is using bash syntax that sh doesn't understand. As you've discovered changing the shebang to #!/bin/bash fixes the problem. You should submit a bug report and hopefully the maintainers will fix it to be less Linux centric.


3

I used the ipadm command. Process documented here http://docs.oracle.com/cd/E19963-01/html/821-1458/gjwiq.html


3

When I run format, and select disk 0, I'm told c4t0d0s3 is part of the root pool: This would be because you've screwed up the slice - it is overlapping with c4t0d0s0. You would need a slice starting at cylinder 3913 to remove the overlap. Then you should be able to add it as a vdev to another pool (if this is what you are after). If you just want to ...


2

First, can you show the output of zpool upgrade -v? Then see: Can I trust that ZFS is consistent between Linux and FreeBSD? The pool version is 28, but "Feature Flags" have been implemented to distinguish functionality between variants of ZFS on different platforms. For instance, a zpool get all vol1 on one of my ZFS on Linux systems ends with: vol1 ...


2

The easiest way is to use pkgrecv to pull the package from a source repository to a local repository. IPS is a network package system so you can't currently install packages from a single file like you could with SVr4 datastreams. Last I heard (before the Oracle takeover) was that the pkg5 guys were working on an on-disk format but I don't think it got ...


2

On question #1: ZFS boot pools can only be single disks or simple mirrors. You cannot use multiple mirrors or RAID-Z pools for booting at this time, due to limitations in boot loaders (GRUB on x86, OpenBoot on SPARC). IIRC, there is a Google Summer of Code proposal to get GRUB 2.0 support into OI, which might allow for booting from more complex pools. For ...


2

about 1: Install OpenIndiana on a single disk (mirroring is not supported by the base installer, if you need, you have to do later) 2. Create your pool when OpenIndiana is running 3. If you use napp-it, you can create pools with the pool-menu and share folders via the share menu (just klick on a ZFS-folder - share smb column) among other management tasks ...


2

I wrote an OpenIndiana tutorial a while back. The other thing it has is a simple explanation of how to kind of lock it down a little. I have some friends and family on the server so I have barely more trust in them than strangers-- in regards to my system. They are still 'users' after all. You can never trust 'users'. ZFS File Server Walkthrough -- EDIT ...


2

Adding to slashdot's comments above: this one-liner does the same thing over all ZFS datasets (as determined by the pattern of the device column in mount(1M) output: fsstat -i `mount | awk '{if($3 ~ /^[^\/:]+\//) {print $1;}}'` 1


2

I experienced this very same issue with a special request to have a service provider install Nexenta on their storage platform. The problem is OpenSolaris/OpenIndiana generally only runs on older hardware because it forked a long time ago and there is a limited selection of drivers. A newer version of Solaris may still not help because the same issue of a ...


1

Don't use ifconfig to plumb the interface use ipadm create-if instead. You've also managed to configure the adaptor 9 times, all of which are configured to use DHCP. You should run ipadm delete-addr on all 9 and perhaps even ipadm delete-if vmxnet3. Once you've done that, you should also stop using vmxnet3 interfaces in your virtual machine configuration ...


1

OK, it's clear now. That happens because zpool list reports the total size and free space of all your storage space, without taking into account that you asked for redundancy. In fact, 22*3*1000^4/1024^4 amounts to 60.0, which sounds close enough to 59.8 (I don't know the exact size of the 3 TB disks). On the contrary, zfs list (and the screenshot you ...



Only top voted, non community-wiki answers of a minimum length are eligible