Recent content by Jerkson

I generally do this for my client's so that I have a backup to restore from when formatting. However, most of my client's like that they have a backup of everything kept from that point in time in case of disaster. I purge them every 6 months or so and when they are stored they are encrypted...

I figured it was better to revive this thread than to create a new one....so here goes. Performance over the last week has been pretty good at 80MB/sec transfers and 340MB/sec resilvers (replaced the 3 Seagates with Hitachi's) I checked and all my drives are at firmware A180 as they have...

I'm not going to argue but it doesn't really matter if you had 100 of these, if they were all AFTER the time/batches where they were setting them to 1.5 this would not pose an issue to you. My drives were manufactured in Dec 2010. I will plug them in directly again tonight and check. The Hitachi...

You can't say that definitively. This was apparently the case with certain batches of drives. I will connect them direct to my lsi and see what they negotiate at

Thanks for the info. That's extra strange that one works and the other doesn't for you. I was wondering if there were some settings that should be changed on the LSI HBA using lsutil to assist in proper negotiation. My concern is that if I add an SSD for a read cache that it's going to...

Im using the HP expander with an LSI-9211-8i (SAS2008) HBA in IT mode. All my Hitachi 7k3000 2TB drives are negotiating at 1.5gbps as per LSIutil: Main menu, select an option: [1-99 or e/p/w or 0 to quit] 69 Seg/Bus/Dev/Fun Board Name Board Assembly Board Tracer 0 3 0 0...

Im sorry but what version of ESXi? Starting with version 5 they do the ACS switch checking. I see you did a grep on vmkernel.log for ACS Try to passthru, reboot as it says you need to to activate the device for passthru, then check vmkernel.log for "non-acs" I experienced the same thing and...

Well it looks like these drives didn't lie well enough ;) The pool is set to ashift=12. From the reading I have done using ashift =12 (even on non-4k drives_ would be preferred especially if there is a chance you would use 4k drives in the future. There should be no performance penalty from...

Every config I tested yields 110MB/sec? I'm assuming you mean in dd tests... as _Gea said above regarding dd: Indeed local dd tests are all well above 110MB/sec. "zdb datapool1 | grep ashift" returns ashift=12. Shouldn't this be 9 with non 4k drives? How is this number automatically chosen...

Here's an update from my benchmarks in case anyone is interested. I know they are not scientific or anything but it gives me an idea of the performance difference between different pool configs...all other things being equal. Copying a ~10GB MKV video from a Win 7 PC to SAn through a SMB share...

Perhaps I will do a comparison with a zpool of 2 raidz2 (5x2tb) just to see if throughput is better everything else remaining constant. If not, I will re-do the 11 drive raidz3 and start adding ram. I will also passthru a physical Intel nic to the VM to see if that helps any.

Good to know about the NFS. The Sync property of the folder is set to "standard" as are the other folders. Is this what should be turned off? I read about NOT doing that... Using an SSD as Read Cache wont speed up writes, will the RAM speed up both directions? If this is the fastest it will...

I've been running tests throughout the day and copying via NFS from windows is painfully slow @ 12MB/sec. At least SMB was quicker. Still trying to sort out why it's so slow.

Any ideas? I dont want to just start tuning at random..

So here's where I'm at with all this. I started fresh to do a good test before deploying. Setup: Spun up the new VM in Esxi4.1: 6gb RAM, 40gb disk, 2vcpu @ 2.0ghz, Fresh install of OI updated to 151a3. Install VMware tools PCI-Passthrou'd the sas2008-based 9211-8i in IT mode with 11x2TB...