Slow E3500 disk throughput in Solaris 8?

I am syncing two SUN72G disks using Disksuite and getting only about 4.7MBytes/sec throughput (2.3 or of read on one and 2.3 write on the other). Seems like it ought to be faster.
This is a 3500 running Solaris 8. I have another server and can get about 7.5 MBytes/sec during a Disksuite resync.
I don't see any Socal. Sf, etc patches for Solaris 8.
Is this normal for these boxes?

Actually this probably doesn't have anything to do with the disk errors, which seems to be its own problem. The difference between the two systems, the slow on and fast one, is that I started a second metattach on the same pair of disks. With one metattach running, I get 24+ MBytes/sec, with two metattaches running on the same pair of disks I get 2.3MBytes/sec.
I know I get a performance hit for doing this, I see it happen in SCSI but not a reduction to 1/10th of normal speed. This must be some kind of evil interaction between Disksuite, socal, and sf.
Message was edited by:
wsanders

Similar Messages

Disk size in Solaris 10

I have some confusion about disk subsystem in Solaris, i am trying to clarify from this forum.
I have recently installed Solaris 10 in one SPARC box. After i installed, the format gives the bellow output.
0 root wm 19491 - 29648 4.88GB (10158/0/0) 10239264
1 swap wu 0 - 4062 1.95GB (4063/0/0) 4095504
2 backup wm 0 - 29648 14.25GB (29649/0/0) 29886192
From the above output, is the size of my disk is 14 GB ?, or the size of my disk is 14+2+5=21 GB ?
I am trying to learn ZFS, so i want another partition in this disk so that i create ZFS on that partition.
I have gone to single user mode by using CD. I assumed that, from the above "format" command output, i thought i have 21GB of disk size and 14GB of free space. So i created another partition with 14GB. Now the format command gives bellow output.
0 root wm 19491 - 29648 4.88GB (10158/0/0) 10239264
1 swap wu 0 - 4062 1.95GB (4063/0/0) 4095504
2 backup wm 0 - 29648 14.25GB (29649/0/0) 29886192
3 reserved wm 0 - 29127 14.00GB (29128/0/0) 29361024
When i am creating ZFS, it given me a warning that the the partition i have specified is spanned into root partition (first partition), and it mentioned to use "-f" option.
With "-f", it created successfully.
If i assume now that the size of my disk is 14GB only then,
(1) how come two partitions are pointing to the same area in the disk ?
(2) How come two different filesystems are pointing to the same area ?
Please anyone clarify my doubts. Thank you.

Assuming a standard labeled disk it is standrad practice to have section/slice 2 being 'whole disk' for purposes of 'backup'. That would tend to indicate you have a 14GB disk. A prtvtoc /dev/dsk/c?t?d?s2 (change the ?s to the right values) will give a little more on the disk geometry.
In the display from format column 4 is the start cylinder of the partition and column 5 is the end cylinder. From the first set out output it looks like cyclinders 4063 to 19490 are not allocated
In the second set you have created a new slice (section 3) that overlaps both sections 0 and 1 - which is generally considered to be bad!

Windows 2008 R2 Hyper-V has really slow VMs disk access

Wondering if anyone can help? I have noticed in a couple of Windows Server 2008 R2 Hyper-V scenarios, that the VM guests are often very slow. Disk access is especially slow.
In one situation their was only a couple of guest VMs, and they were running of mirrored IDE HDDs. Performance improved significantly when I used two mirrored SSD hdds instead.
My other main situation is where the VHDs are stored on HP Lefthand P4300, over 2-nodes with 8 SATA hdds in each, using iSCSI.
So I ran a few tests with robocopy copying 4-6GB of files (mostly large) or whole VHDs (40GB or more), using HP4300 SATA, a newer HP 4330 SAS, a THECUS NAS with SATA hdds, and an SSD in a THECUS NAS, and local SAS mirrored hdds. All using iSCSI except
the local SAS hdds.
You can see from results that Guests run considerably slower than host. Any ideas?
Sample data (although not always same background conditions of load etc):
From within Guest direct iSCSI to same (from HP 4330): 34MB/s
From other Guest using VHD (from HP4300) to direct iSCSI (from HP4330): 14MB/s
From within Guest using VHD (from HP4330) to same: 9MB/s
From within Guest using VHD (from HP4300) to VHD (from HP4330): 11MB/s
From within Guest using VHD (from HP4330) to VHD (from HP4330): 19MB/s
From within Guest using VHD (from THECUS SSD) to same VHD (from THECUS SSD): 18MB/s
From within Guest using VHD (from HP4300) to VHD (from THECUS SSD): 13MB/s
From within Host from HP4300 iSCSI to HP4330 iSCSI: 12MB/s
From within Host using VHD (from THECUS SSD) to same VHD (from THECUS SSD): 40MB/s
From within Host from HP4330 iSCSI to same HP4330 iSCSI: 232MB/s & 132MB/s
From within Host from HP4330 iSCSI to HP4300 iSCSI: 40MB/s & 57MB/s
From within Host from HP4300 iSCSI to same HP4300 iSCSI: 26MB/s & 47MB/s
From within Host from HP4300 iSCSI to Guest VHD (from THECUS SATA): 15MB/s

Hi aucj,
I would suggest to use host level ISCSI connection then build VM on that ISCSI disk .
please add more Vcpu for that VM then try to copy again (shutdown the other VMs ,in that online VM from one folder to another folder ).
Hp4330 ISCSI connection is 1GB , right ?
If yes , it should be with max 100M/s with the single ISCSI connection .
Best Regards
Elton Ji
We
are trying to better understand customer views on social support experience, so your participation in this
interview project would be greatly appreciated if you have time.
Thanks for helping make community forums a great place.

How can I get the disk information in Solaris 9

I am new to Solairs9. I use the format utility to list my partitions, but there are many unassigned partitions. I pasted the format response below:
# format
AVAILABLE DISK SELECTIONS:
0. c1t1d0 <SUN72G cyl 14087 alt 2 hd 24 sec 424>
/pci@8,600000/SUNW,qlc@4/fp@0,0/ssd@w21000011c6413e10,0
Specify disk (enter its number): 0
format> partition
partition> print
Current partition table (original):
Total disk cylinders available: 14087 + 2 (reserved cylinders)
Part Tag Flag Cylinders Size Blocks
0 root wm 302 - 13482 63.96GB (13181/0/0) 134129856
1 swap wu 0 - 301 1.47GB (302/0/0) 3073152
2 backup wm 0 - 14086 68.35GB (14087/0/0) 143349312
3 unassigned wm 0 0 (0/0/0) 0
4 unassigned wm 0 0 (0/0/0) 0
5 unassigned wm 0 0 (0/0/0) 0
6 unassigned wm 0 0 (0/0/0) 0
7 home wm 13483 - 14086 2.93GB (604/0/0) 6146304
The part number from 3 to 6 are unassigned. What does this mean? How much space of these unassigned partitions? How can I mount these unassigned partitions to my local directories?
Thanks.

"+How can I get the disk information in Solaris 9+ "
This was cross-posted to the Solaris 9 forum as well.
http://forum.java.sun.com/thread.jspa?threadID=5223643

Setup Disk Mirroring on Solaris 8

I need to know of how to setup Disk Mirroring on Solaris 8 on my SUN E250/E280
Which manual collection, provide a step by step procedure to do it. Thanks.

I assuming you're going to be using Sun disc management software ( DiskSuite) and not something like Veritas volume manager.
There are some Sun infodocs on the sunsolve site ( sunsolve.sun.co.uk is the one I use ) that contain good step by step guides on how to mirror discs using Solstice Disksuite. Type in "disksuite faq" as keywords to search.
Disksuite 4.2.1 comes with Solaris 8 ( this is the recommended version of DiskSuite you use with 8), but you need to make sure it was included as a package to install when the server was originally built). If it hasn't been installed, put the Solaris 8 cd 2 of 2 and do a package add on the following path /cdrom/sol_8_1000_sparc_2/Solaris_8/EA/Products.
Pick disksuite and I generally add all the packages. You'll need to know the product a bit more if you want to pick and choose the components to install.
Once done, you'll need to reboot the server and then begin to configure the discs. To understand the various concepts with disksuite you will need to read the manuals.
I use the command line but there is a GUI tool supplied with the product called metatool to help you mirror your discs.
As a high level overview this is what you need to do:
Create a minimum of 3 state databases ( on a separate disc slice of about 5 megabytes in size, nothing else should use this slice )
The discs you want to mirror should have the same partition table to keep life simple.
I would create 6 databases, 3 on an individual slice.
Then create your mirror devices which represent the various partitions you want to mirror from the primary disk. Then create your metadevices and attach the first set of mirrors to them.
You need to issue a special command called metaroot to get the system to mount the root filesystem metadevice at boot time.
At this stage I would then make a backup of /etc/vfstab and then edit it to mount your metadevices instead of the normal disc devices.
I would then bounce the server to check it all works and the metadevices get mounted correctly.
Create your second set of mirror devices on the secondary disc and add them to the appropriate metadevice ( using metattach )
If the filesystems are very large it takes a few minutes for the disks to sync up, so don't bounce the server until this process has completed. You can run a check on disk synchronisation by doing metastat | grep sync
Finally I would put the output from the command metastat -p into the /etc/lvm/md.tab file at the very end with some appropriate comments.
This can seem a very onerous task at the beginning but once your clear on steps you need to run, the whole procedure is quite straightforware.

Add/Remove Raw disk to a Solaris 11 Container

Hi All,
What is the process or commands used to add or remove raw disks to a Solaris 11 containers.
I need to present the raw disks to containers so that Oracle ASM can be used on the containers. I have added the raw disks but not able to delete the extra disk. Can somebody help on this. Thanks in Advance.

I've not tried this with disk (only with tapes and that years ago on Solaris 10) but I think you need something add something like the following to your zone's configuration using zonecfg.
add device
set match=/dev/rdsk/<a pattern which matches the disks you want to import>
end
See the Solaris 11 zonecfg manual page (example 8 applies).
Again - I've not tried this so I can't swear that it will work.

How to find a unused hard disk in Sun Solaris unix

Hi All,
Can any one tell, how to find all the unused hard disk in Sun Solaris using Unix command.
Thanks & Regds
PAN

not understand your question, find command is used to look for files, not disks.
easier way is to logon to your system open up 2 windows, on 1 window # cat /etc/vfstab
on second window type format.
format will list all the disks in your system.
vfstab file will list all the potential disks and slices used as swap devices and filesystems like your os and application filesystems etc ...
then you compares left and right.
but even then be very very very careful if you find disks in "format" not found in /etc/vfstab doesn't mean disk not in use, if your system is clustered the clustered resources like mounted volumes will not show up under the /etc/vfstab
if you not familiar with unix u better dun touch.

Disk suite for Solaris 7 and 6

Hello
Can someone tell me where to find the latest version of Disk-suite for Solaris 7 and 6.
Thanks

Yes, you can install on your new machine.
Mylenium

Slow FTP throughput using solaris 9 and max window size question

I am working on a solution where we are trying to FTP at a high rate across the WAN on OC-12 links. We have been doing some TCP tuning with consideration to the Bandwidth Delay Product. Ping results in a 90ms RTT.
I have been told that we will not see any increase in performance above window sizes of 64k without certain patches for Solaris 9. I know that high speed transfers are possible if tuned up correctly with other operating systems.
Does anyone have any knowledge or experiences that they could share with me?
Thanks

You will have to format your disk with the help of the WRT350N and after that you will have creat Share with same "Storage" tab on the linksys router.

Mac mini running slow, will disk utility in machine be effective

Mac mini came installed with 10.4 when purchase long time ago. Looks like the dvd reader no longer works since any cd or dvd gets ejected. The mini has been running slow and I cannot restart with Mac Install DVD to run the Disk Utility since it gets ejected. Will the Disk Utility that is found in machines Utilities be just as good for repairing compared to the Install DVD? Current OS is 10.6.8.

Try it. Some disk repair requires booting from other than the disk being repaired.

Re-read disk size in Solaris 10 x86 / VMware

Hi
Using
# uname -a
SunOS host1 5.10 Generic_142910-17 i86pc i386 i86pc
We have a 30G virtual disk, on top of which a UFS filesystem is living, mounted, and used.
The Virtual disk has been resized to 40G.
I know I need to change the fdisk partition size, after which i can change the "format" partition size, then use growfs.
But I don't know how to get S10 to recognise the new size for the disk.
fdisk tells me
Total disk size is 3916 cylinders
Cylinder size is 16065 (512 byte) blocks
Cylinders
Partition   Status
Type
Start   End   Length
=========   ======
============ =====   ===   ======   ===
1
Active
Solaris2
1 3915
3915
100
# prtvtoc -h /dev/dsk/c1t2d0s2
       0      0    00      16065 62846280 62862344   /xxx
       2      5    01          0 62862345 62862344
       8      1    01          0     16065     16064
# df -h /xxx
Filesystem             size   used avail capacity Mounted on
/dev/dsk/c1t2d0s0       30G    29G   498M    99%    /xxx
On Linux I might use partprobe or blockdev --rereadpt but on Solaris?
Cheers
KM

Hello
this document should help, but make a backkup before running it
How to expand a UFS Lun in Solaris 10 (Doc ID 1451858.1)
Regards
Eze

Disk not found - solaris 10 installation

Hello *!
First of all: thank you for taking the time.
The problem: trying to install solaris 10 on a Enterprise 220R with a D1000 Storage Array attached. The configuration, as seen in OB, looks like this:
/pci@1f,4000/scsi@2
Target f
Unit 0 Proc Symbios D1000
/pci@1f,4000/scsi@3
Target 0
Unit 0 Disk Seagate SX318203LC
Target 6
Unit 0 CD Toshiba DVD-ROM
the alias for disk is /pci@1f,4000/scsi@3/disk@0,0
the probe list entry looks like:
pcib-probe-list 1,3,2,4,5
When I boot off the cd I get to the start of the solaris installation and after configuring the network parameters the insatll stops because of no disk beeing found. Now, if I insert a disk in the D1000, this one will be recognized and used as the install disk.
I guess it has something to do with the fact, that the controller the D1000 attaches to is listed before the onboard one. So, my question: how can I tell the solaris 10 install program to search also the onboard controler? Is there an obvious parameter I overlooked?
Thank you very much and in advance for any help on this topic!
Best regards,
Christian
elsnerc -at- yahoo -dot- de

I think that the problem is more than likely going to be either a faulty disk, or more likely the wrong path that you have. What you can do is the following:
1. From the ok prompt insert the solaris cd 1 of 2 and type 'boot cdrom -s'
This will bring the system up from the CD using a memory stored kernel that is taken from the CD. Once here you can do your normal checks, format, etc, to check the status and paths for the disks.

How to increase disk size in solaris

Hi all
I have two disk c1t0d0s0 / c1t0d0s7
I do haven't space in disk /dev/dsk/c1t0d0s0
run df -h
/dev/dsk/*c1t0d0s0* 8,8G 7,4G 1,3G 86% /
/devices 0K 0K 0K 0% /devices
ctfs 0K 0K 0K 0% /system/contract
proc 0K 0K 0K 0% /proc
mnttab 0K 0K 0K 0% /etc/mnttab
swap 3,4G 984K 3,4G 1% /etc/svc/volatile
objfs 0K 0K 0K 0% /system/object
sharefs 0K 0K 0K 0% /etc/dfs/sharetab
/usr/lib/libc/libc_hwcap1.so.1
8,8G 7,4G 1,3G 86% /lib/libc.so.1
fd 0K 0K 0K 0% /dev/fd
swap 3,4G 72K 3,4G 1% /tmp
swap 3,4G 40K 3,4G 1% /var/run
/dev/dsk/*c1t0d0s7* 20G 21M 20G 1% /export/home
I want increasing the space in c1t0d0s0. when I installed Solaris 10 I do not looked attentively the size allocated
Thanks

Hi,
As I understand, you want to increase the size of the root partition (where you installed Solaris) which resides in c1t0d0s0. I think it can be done using Concatenations from Solaris Volume Manager (SVM).
I have tried it in non root partition like /export/home and it is very easy.
metainit -f d0 2 1 c1t0d0s7 1 c2t0d0s7 //this will create volume d0 as a concatenation
growfs -M /export/home /dev/md/rdsk/d0 //this will grow the /export/home partition
Regards,
Rei

Disk corruption in Solaris 10 with mirrors

I have had 5 different Sun sparc servers in the past couple of months come up with badly corrupted file systems. I am using They are all mirrored metadisks. One thing I have looked into is using the UFS logging option, but from what I read, it doesn't need to be specified in /etc/vfstab because it is on by default in Solaris 10. Is this true?
Also, is there a problem with shutting down or rebooting the OS with the init commands? Should I be using shutdown instead? Does init 0, 5 or 6 cleanly shutdown the system?
Thanks

Hello,
Init 0 is very nice command which is gracefully shutting down all the services in the system including the OS. But the system will be powered on unless you execute powerdown from rsc.
In the case of mirroring metadisks entries can reduce a lot of file system related errors because those entries ("device to fsck") forces the filesystem check in those meta devices in every reboot.
Please have look on vfstab in my test server where i have mirrored disks.
bash-3.00# more /etc/vfstab
#device device mount FS fsck mount mount
#to mount to fsck point type pass at boot options
fd - /dev/fd fd - no -
/proc - /proc proc - no -
/dev/md/dsk/d1 - - swap - no -
/dev/md/dsk/d0 /dev/md/rdsk/d0 / ufs 1 no -
/dev/md/dsk/d2 /dev/md/rdsk/d2 /var ufs 1 no -
/devices - /devices devfs - no -
ctfs - /system/contract ctfs - no -
objfs - /system/object objfs - no -
swap - /tmp tmpfs - yes -
bash-3.00#
Please do "iostat -En" and check for the errors so that you will get a better understanding.
Thanks,
Sal.

No disks found. Solaris 10.

When i have tried install Solaris 10 on my laptop HP dv 6000 i have received this message "no disks found. chceck to make sure are cabled and powered up." what can i do? Maybe i should mount the disc? But i dont know how, i tried but i had a problem witch this, it is different than mount disk at Linux.

I was able to install Solaris 10 on a hp dv6000 as a dual boot system with Microsoft Vista. After installation I had two problems.
1) GRUB seems to have changed the Disk Signature in the MBR making Vista unbootable. See:
www.multibooters.co.uk/mbr.html.
I found a possible solution at: http://blogs.sun.com/moinakg/entry/vista_and_solaris_express_dual
but haven't tried it yet.
2)Installing the Patches hung the system and made Solaris 10 unbootable. I haven't tracked down which patch was responsible.
Message was edited by:
JodyN
Message was edited by:
JodyN

Slow E3500 disk throughput in Solaris 8?

Similar Messages

Maybe you are looking for