OpenSolaris derived ZFS NAS/ SAN (OmniOS, OpenIndiana, Solaris and napp-it)

the bonnie benchmark on napp-it, SEQ-WRITE speed is talking about write from disk or to disk ? if its write from disk than i know whats wrong already
 
kegler said:
i3-2100T
dell 5i hba
J&w H61 mini itx
dual broadcom nic

oi 148
napp-it 0.5

Intel g2 SSD 80GB (boot
Seagate constellation 1TB x 2 (mirror)
Samsung F4 2TB x 2 (mirror)


the write speed is about 80MB/s but read speed is on 40+ MB/s


I couldnt figure out whats wrong....anyone has any idea?
Click to expand...

i had the same problem the last two weeks :(
google zpool-12 you will find the solution ...

I might be wrong but i think its the same 4k problem i had....
 
Update on the whole network bridge thing....

So I did a clean install of OpenIndiana 151a and everything workedout very smoothly
Set static ip on ixgbe0, create new bridge, add ixgbe0 and ixgbe1 to the bridge. Everything works, hosts on both side of the bridge can communicate.
 
axan said:
Update on the whole network bridge thing....

So I did a clean install of OpenIndiana 151a and everything workedout very smoothly
Set static ip on ixgbe0, create new bridge, add ixgbe0 and ixgbe1 to the bridge. Everything works, hosts on both side of the bridge can communicate.
Click to expand...

Perfect. Thanks for the update!

Out of curiosity, which 10GbE adapters are you using?
 
Gea, I'm trying to work out some ACL issues and when I went to the ACL extension in napp-it I got this:

Folder not allowed - not a datapool -??

any ideas whats wrong?

Thanks!
 
s0rce said:
Gea, I'm trying to work out some ACL issues and when I went to the ACL extension in napp-it I got this:

Folder not allowed - not a datapool -??

any ideas whats wrong?

Thanks!
Click to expand...

you should give some details like OS/ napp-it version and zfs list
if you use 0.6 preview, try menu zfs reload first

(ACL extension is not yet ready, just like napp-it 0.6 preview - less time to work on it)
 
Rectal Prolapse said:
kegler it isn't clear if those speeds are over network or local.
Click to expand...

It's over the network... Giga switch with oi side broadcom nic and pc side realtek

I tried transfer from another pc on the network.... Write 30mb/s read 60mb/s which seems to be more normal..... I seriously can't figure out what's wrong with my oi

Sigh
 
kegler, hard to keep track of who has posted what, so if you've said this already forgive me. do you have sync mode set for the dataset (or the pool itself)?
 
danswartz said:
kegler, hard to keep track of who has posted what, so if you've said this already forgive me. do you have sync mode set for the dataset (or the pool itself)?
Click to expand...

no one asked...thanks for replying :)

sync mode is standard

if disable it, wouldnt it be dangerous ?
 
It depends. As I understand, zfs will never have filesystem corruption, but if the NAS crashes during writes, the client may lose data. The performance difference can be major. Using an SSD for a ZIL helps a lot (it only needs to be 1/2 RAM or so...)
 
I just got a ASRock Z68 Extreme4 Motherboard. It seems the Broadcom BCM57781 Gigabit Ethernet controller is not supported by OpenIndiana or Solaris... Ugh?
 
Need some help. I set up an ESXI 5 deployment, passed my disk controller to an OpenIndiana 151 VM and my disks show up fine. Installed napp it using the all-in-one pdf guide. Created a raidz2 pool, added an NFS folder Then when I try to add the NFS folder as a datastore in ESXI, I get the following message "Unable to get Console path for Mount". Anyone have any clue what's going on? I know the folder path is write, because if I change it, it gives me a bad location error. Thanks for your help!
 
Woke up to a power failure last night and it looks like my UPS kept everything online for ~20 minutes, but it all shut down hard.
Napp-it is giving "Failure" Corrupt Data on all of my drives on my LSI 1068...
:(
 
Power off and back on doesn't find the drives? Or are they really corrupt? What does 'zpool status' show from CLI? Silly question: why isn't your system set to shut down when the UPS says time's up?
 
danswartz said:
Power off and back on doesn't find the drives? Or are they really corrupt? What does 'zpool status' show from CLI? Silly question: why isn't your system set to shut down when the UPS says time's up?
Click to expand...

That's something that I need to figure out. Like the poor guy above, my UPS will just stay on (not the best solution obviously). Since I have an all-in-one, I'm thinking I can create a Windows XP VM and load the UPS software on that. From there, however, I need to find a way for that VM to shutdown ESXi. Anyone else have an easier solution already?
 
unclerunkle said:
That's something that I need to figure out. Like the poor guy above, my UPS will just stay on (not the best solution obviously). Since I have an all-in-one, I'm thinking I can create a Windows XP VM and load the UPS software on that. From there, however, I need to find a way for that VM to shutdown ESXi. Anyone else have an easier solution already?
Click to expand...

I remember reading about UPS solutions with ESXi. I have solaris running with apcupsd (no virtualization) so I don't need to worry but there might be a solution out there.
 
Oh poo, sorry didn't realize it was an all in one. Well, what I'm going to do (when I get time) is to set apcupsd to run on the OI VM and send a shutdown command to the hypervisor. It's easy to get apcupsd working on OI, although you need to compile it from scratch. Actually, I have a separate centos box for my ups, and I have apcupsd on OI monitoring the UPS via the network (e.g. it connects to the apcupsd on the pbx via tcp/ip). I seem to recall that OI 148 USB didn't work well for me - it would fault and die (a USB driver issue, if I recall...)
 


Uploaded with ImageShack.us

Yep, it is a all in one and it looks like I will be learning the same thing danswartz... I want to use a similar setup in a production environment and this just wont due. I need to rebuild the pool anyways for my additional drives and SSD cache, but I really, really hate to do it because of this... The only thing I can think of is I rebuilt the pool 3 times learning napp-it, but I have not seen anyone else loose all their drives like this.
 
Cool, I was hoping that might work for you. If you want, you can rename the pool back to the old name...
 
Hi S0urce/Dan, re. UPS shutdown for an all in one, I thought you mat find my setup of interest.

I have an "all in one" server running of an APC UPS. The Solaris 11 Express vm is setup under ESXi 4.1 serving as a SAN for a number of esxi vm's. It is the first to boot and last to shutdown. I have installed apcupsd on this S11E machine (vm) and passed through the serial port from the APC UPS...(you could just as easily pass the usb port from the ups if that is what your uses).

apcupsd monitors the ups. If it needs to shutdown due to power loss, it runs a script to ssh into ESXi and send the shutdown command to the other vm's and then a "general" shutdown command to ESXi which also shuts down the S11E machine and the whole server.

You can read about an example of how I setup the ssh in my post here

You can find the esxi command line commands on the web...they give you great control of the vm's...

eg, to shutdown the vm with the machine id of "600", where the ip address of the esxi box is 192.168.0.31, the command on the S11E machine would look like this.
Code: 
ssh 192.168.0.31 -l root "vin-cmd vmsvc/power.shutdown 600"

I hope that helps and keen to hear if thereis a more egant way to handle this.
 
Davros, that's basically what I was going to do with one change: ssh in and run /bin/shutdown.sh. That is a shell script on the hypervisor that shuts down all the autostart VMs (so I don't have to manually.) It isn't obvious/clear that it powers off the host itself, but if the VMs are all down, I don't care I guess...
 
I'm running S11E on an AMD Athlon II X2 250 and I was wondering how to enable CPU throttling to save power/reduce heat load.

Code: 
kstat -m cpu_info -i 0
module: cpu_info                        instance: 0
name:   cpu_info0                       class:    misc
        brand                           AMD Athlon(tm) II X2 250 Processor
        cache_id                        0
        chip_id                         0
        clock_MHz                       3013
        clog_id                         0
        core_id                         0
        cpu_type                        i386
        crtime                          52.90959789
        current_clock_Hz                3000000000
        current_cstate                  1
        family                          16
        fpu_type                        i387 compatible
        implementation                  x86 (chipid 0x0 AuthenticAMD 100F62 family 16 model 6 step 2 clock 3013 MHz)
        model                           6
        ncore_per_chip                  2
        ncpu_per_chip                   2
        pg_id                           2
        pkg_core_id                     0
        snaptime                        192856.966727097
        socket_type                     AM3
        state                           on-line
        state_begin                     1316380034
        stepping                        2
        supported_frequencies_Hz        800000000:1800000000:2300000000:3000000000
        supported_max_cstates           0
        vendor_id                       AuthenticAMD

It appears that it knows the CPU supports lower freq, down to 800MHz but the system is idle now but still is at 3GHz.

Any help would be great. I didn't find anything in the bios that appeared to have disabled throttling.

Thanks!

edit:

I added:

Code: 
autopm                  enable
autoS3                  enable
cpupm                   enable poll-mode
cpu-threshold           1s
cpu_deep_idle           enable
loadaverage             0.07

to my /etc/power.conf and then ran pmconfig. It appears to now throttle the CPU, I'll have to check the bios at some point to ensure c-states are enabled.
 
Last edited:
Quote:
Originally Posted by danswartz View Post
Here is what works on my OI install with sata drives:

smartctl -a -d scsi -T permissive /dev/rdsk/c3t0d0
Thanks you, always wondered how to check the temp of my drives.
Click to expand...

Tried this command, however I get

Code: 
smartctl 5.40 2010-10-16 r3189 [i386-pc-solaris2.11] (local build)
Copyright (C) 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net

Smartctl open device: /dev/rdsk/c3t0d0 failed: No such device

Any guide on how to get the temperature of the hard drives to display and how to add it to the napp-it menu?
 
Here are my 6 data pool drives, fetched as described:

Current Temperature: 37 Celsius
Current Temperature: 41 Celsius
Current Temperature: 40 Celsius
Current Temperature: 37 Celsius
Current Temperature: 40 Celsius
Current Temperature: 38 Celsius

I run the command for each disk and pipe into:

grep 'Current Temperature'
 
I'm having problems with Comstar Fibrechannel and iSCSI setup. The FC works flawlessly but i cannot add any working LUs to iSCSI for some odd reason :(
I'd really like to recover my steam from there.

When i add new lun it shows as it should in napp-it and i can assing it to views and all but no. iSCSI iniatators only see the FC LUN :|
 
Never mind, got it running by removing and recreating the lun and setting it's views to all. Suprisingly the data is intact :p
 
danswartz said:
You need to put your actual device name(s) in instead of c3t0d0 :)
Click to expand...

Duh, :) However still getting errors... any ideas?

smartctl -a -d scsi -T permissive /dev/rdsk/c0t50024E9004DFF341d0

smartctl 5.40 2010-10-16 r3189 [i386-pc-solaris2.11] (local build) Copyright (C) 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net Serial number: DDDDDDDDDXXXX Device type: disk Local Time is: Thu Sep 22 03:41:56 2011 CDT Device supports SMART and is Enabled Temperature Warning Disabled or Not Supported Log Sense failed, IE page [scsi response fails sanity test] Error Counter logging not supported [GLTSD (Global Logging Target Save Disable) set. Enable Save with '-S on'] No self-tests have been logged Long (extended) Self Test duration: 20400 seconds [340.0 minutes]
 
The components of my All-in-One are arriving. I will be creating a 6x2TB raidz2 ppol using Hitachi 512k drives. In the mean time, I have created a VM running OI 151a and napp-it to begin learning.

Question 1: What is the most stable release of OI that I should use?
Question 2: If Solaris Express 11 is more stable than OI, would you recommend SE11 (I have no need for encryption).
Question 3. In my test VM, I created a 6 drive raidz2 pool, however, I can't tell if the pool that was created is actually a raidz2 or raidz pool (the pool status differs from the pool details).

922201134028pm.png
 
Last edited:
You must log in or register to reply here.
Back
Top