haimg's questions

A VMWare ESXi 6.0 server used to be controlled via a serial console, but now has IPMI card with VGA redirection. During the ESXi boot process, there's a running list of "Loading /xxxx.tgz", then "Relocating modules and startup the kernel". Nothing appears on the (network redirected) screen after that.

I've set VMkernel.Boot.tty2Port to "none", and to "default", but that made no difference. All the relevant services are loaded, and running. Nothing relevant is in the logs.

How can I get my local console back?

Edit: the server runs fine otherwise, just the local console is not available.

I have VMWare ESXi 5.1 host with several gigabit links, connected to a LACP and 802.3ad-capable switch, in a static LACP mode on a switch, with "IP hash" load balancing mode in ESXi. I'm currently using vSphere standard switch.

It seems like while load balancing works ok for multiple TCP/IP streams to/from guest VMs, no single stream can utilize more than one link (e.g. 1Gbps), due to IP hash load balancing. And this is of course the only load balancing mode that supports LACP on ESXi.

Is there anything I can do to allow one TCP/IP stream to/from guest VM to utilize several gigabit links at the same time?

I'm creating RAID10 array from 6 drives. When created in near layout, e.g.

mdadm --create /dev/md2 --chunk=64 --level=10 --raid-devices=6 --layout=n2 /dev/sda1 ...

Checking stripe size as reported by the system:

cat /sys/devices/virtual/block/md2/queue/optimal_io_size

Result is 196608, as expected, e.g. 3 data drives (50% of 6 total in RAID10) x 64K chunk = 192K stripe.

Now, when creating same array with --layout=f2 option, optimal_io_size reports 393216, e.g. twice as large.

Now, according to Nail Brown (mdadm raid10 author),

The "far" layout lays all the data out in a raid0 like arrangement over the first half of all drives, and then a second copy in a similar layout over the second half of all drives - making sure that all copies of a block are on different drives.

This would be expected to yield read performance which is similar to raid0 over the full number of drives, but write performance that is substantially poorer as there will be more seeking of the drive heads.

So it seems OS is suggesting I'd better use RAID0-like stripe size (across all disks in the array), and not the "traditional" RAID10 stripe size (across half disks in the array). This has potentially serious implications for LVM and filesystem alignment, stripe/stride tuning, etc. However, I've never seen anywhere suggestions to treat MDADM RAID10 in far mode as RAID0.

Question: Am I missing something here, or am I correct to treat RAID10,f2 as RAID0 when aligning/tuning whatever lays on top of that RAID?

Our internal network and internal services are IPv6-enabled, and we're considering enabling IPv6 on public-facing SMTP, DNS and Web servers. All the guides / whitepapers I've read on the subject speak only about the benefits, but I'm sure there are also some downsides. For example:

1. Clients with IPv6 via tunnels will have slower access to our services if we enable IPv6.
2. Clients with broken IPv6 won't be able to access them at all.

Given that only very small percentage of users have working IPv6, and those that do are usually dual-stack, I'm inclined to think that enabling IPv6 when you don't control both sides of the communication channel is actually a bad idea in 2011.

Is there any research / papers / articles on this subject?