All USB Drives are not equal?

[beng]

G'day All,
We are having some strange issues with a couple of DSS Lite installations.
In all we have 5 of these in various systems, running as dev or backups.
In two cases, both completely different architectures, eg Intel Server S5000PAL board and a Tyan Thunder (I think) AMD, the DSS Lite is very flaky.

They often just stop responding, ie the console locks up as well, or sometimes you can browse the network SMB shares, but not get into the Web GUI.
But in all cases they still respond to a Ping.....

A couple of times after rebooting, the USB disks also need to be recreated.
So this is making me think that the USB disks may be at fault, one is a no-name 1GB and the other is a 2GB Toshiba. Both were bought at different times.

Does anyone have any suggestions for testing these (and future) usb disks, to ensure they are ok?

Rgds Ben.

[To-M]

On the other 3 DSS LITE boxes are they demonstrating the same problems?

Was all the USB disks formatted as FAT or FAT32?

We have the S5000PAL motherboard (Intel SRS2500) with 1x Xeon 3.73GHz, Bios Version s5000.86b.10.00.0084 and BIOS Build Date 10/17/2007 w/ 2GB Memory FB ECC but tested with iSCSI only not with SMB. We could try but don’t think it is a OS issue for now.

Any chance you can check the logs from the GUI from Status > Hardware > Logs then click on the "Download" button (try with Firefox as IE will save this as a PHP file).

Then open and search for the log file of dmesg.log and post them. I want to see if there could be some hardware issues in the background.

[beng]

G'day Todd,
Hmmm very odd, it could be the SAS/Sata card,
The last few entries are below, and so it does look like the Adaptec Raid Controller may be the issue, but we have also had this happen on an Areca 1120(I think it is)...
Also, if the card/disks are at issue, shouldn't the OS be independent of this?
ie why is it failing to load the Web GUI etc.....
It's not that unusual to get faults in SATA class drives, so I would expect this to be a pretty common issue?

Ethernet Channel Bonding Driver: v3.0.3 (March 23, 2006)
bonding: MII link monitoring set to 50 ms
bonding: bond0 is being deleted...
e1000: eth1: e1000_watchdog_task: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
program scsiinfo is using a deprecated SCSI ioctl, please convert it to SG_IO
program scsiinfo is using a deprecated SCSI ioctl, please convert it to SG_IO
program scsiinfo is using a deprecated SCSI ioctl, please convert it to SG_IO
program scsiinfo is using a deprecated SCSI ioctl, please convert it to SG_IO
kjournald starting. Commit interval 5 seconds
EXT3 FS on dm-13, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
drbd: initialised. Version: 0.7.22 (api:79/proto:74)
drbd: SVN Revision: 2575M build by root@develsystem2, 2007-12-06 08:05:37
drbd: registered as block device major 147
kjournald starting. Commit interval 5 seconds
EXT3 FS on ram15, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
program scsiinfo is using a deprecated SCSI ioctl, please convert it to SG_IO
program scsiinfo is using a deprecated SCSI ioctl, please convert it to SG_IO
kjournald starting. Commit interval 5 seconds
EXT3 FS on dm-7, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
Filesystem "dm-10": Disabling barriers, not supported by the underlying device
XFS mounting filesystem dm-10
Ending clean XFS mount for filesystem: dm-10
Filesystem "dm-11": Disabling barriers, not supported by the underlying device
XFS mounting filesystem dm-11
Ending clean XFS mount for filesystem: dm-11
Filesystem "dm-12": Disabling barriers, not supported by the underlying device
XFS mounting filesystem dm-12
Ending clean XFS mount for filesystem: dm-12
Adding 4194296k swap on /dev/vg+vg00/nas_swap. Priority:-1 extents:1 across:4194296k
NET: Registered protocol family 10
ADDRCONF(NETDEV_UP): eth0: link is not ready
IPv6 over IPv4 tunneling driver
scst: Attached SCSI target mid-level at scsi0, channel 0, id 0, lun 0, type 0
scst: Attached SCSI target mid-level at scsi1, channel 0, id 0, lun 0, type 0
scst: Attached SCSI target mid-level at scsi4, channel 0, id 0, lun 0, type 0
scst: Attached SCSI target mid-level at scsi4, channel 1, id 0, lun 0, type 0
scst: Attached SCSI target mid-level at scsi4, channel 1, id 4, lun 0, type 0
scst: Attached SCSI target mid-level at scsi4, channel 1, id 5, lun 0, type 0
scst: Attached SCSI target mid-level at scsi4, channel 1, id 6, lun 0, type 0
scst: Attached SCSI target mid-level at scsi4, channel 1, id 7, lun 0, type 0
scst: SCST version 0.9.5.2 loaded successfully (max mem for commands 1010 Mb)
scst: Device handler disk_fileio for type 0 loaded successfully
scst: Device handler cdrom_fileio for type 5 loaded successfully
eth1: no IPv6 routers present
Fusion MPT base driver 3.03.09
Copyright (c) 1999-2005 LSI Logic Corporation
Fusion MPT misc device (ioctl) driver 3.03.09
mptctl: Registered with Fusion MPT base driver
mptctl: /dev/mptctl @ (major,minor=10,220)
iSCSI Enterprise Target Software - version 0.4.15
iotype_init(92) register fileio
iotype_init(92) register blockio
iotype_init(92) register nullio
aacraid: Host adapter reset request. SCSI hang ?
aacraid: Host adapter reset request. SCSI hang ?
aacraid: Host adapter reset request. SCSI hang ?

[To-M]

If possible can you test without this card and use a single drive for the Volume Group to see if the GUI still hangs or has issues. In the professional versions of DSS we have other tuning options that I would be able to assist you on the DSS LITE is limited on these options.

[Shankyrhodes]

Hi,

we are considering to buy a VMware Virtual Center.
We have two servers running VMware Standard edition.
Do you believe it will be worth it? Or do we have to
upgrade our VMware licenses to Enterprise before upgrading
virtual center to make it worth it. I had just read the
following article

http://www.virtualizationteam.com/virtualization-vmware/vmware-virtual-server-virtualization-vmware/virtualcenter-for-vm-ware-server-real-value.html

[To-M]

I am not a VMware expert and maybe someone else can provide a better answer but if you are thinking of using the VMotion, HA and DRS then the VI is most likely your best bet.

[Harvin]

Hi,

we are considering to buy a VMware Virtual Center.
We have two servers running VMware Standard edition.
Do you believe it will be worth it? Or do we have to
upgrade our VMware licenses to Enterprise before upgrading
virtual center to make it worth it.

It would be nice to have but expensive for your setup vs feature set. With Std Ed you can not get the features that really make it worth while. For example VMotion requires Enterprise Ed and a SAN. Take a look at the free versions of FASTSCP and RootAccess at www.veeam.com. They will allow you to copy (non-running) VM's. You use that to to simulate templates in VC and copy them between the two servers. I think that duplicates the biggest things you would get from VC in your setup.

[beng]

G'day Shankyrhodes,
Harvin is correct that VC is expensive for what it does.

What features are you looking for?

If you are able to shutdown VM's for short periods while you work on the ESX, then an OpenE iSCSI is all you need.
So long as the VM is on the shared iSCSI target, you just shutdown the VM, remove it from the Inventory, browse the datastore on the other ESX, and import the vmx config.

VC would allow you to cold migrate (ie shutdown) VMS from one ESX to the other if you have a shared data store (ie DSS iSCSI) and it is also nice for templates. (Though Harvins solution is also fine if you don't do them often).
It also gives you longer trending for performance, and I think that would be abuot it in your scenario.

Rgds Ben

[beng]

Just for everyones FYI, the original issue we had above never did get fully resolved with DSS Lite, but the same hardware works a treat under full DSS.
Rgds BEn

[enealDC]

Shank, have you considered Virtual Iron? You get pratically all the features of VMWare at like 1/4 of the price.