A bunch of stuff broke this month, learned a lot fixing it all

Posted by Paul Braren on Jun 18 2014 in
  • ESXi
  • Network
  • Productivity
  • KVM over IP, LSI RAID, VMFS corruption averted, unplanned smartphone replacement, cablemodem replacement, and phone line replacement. What a month! Follow along and learn from my mistakes, and successes.

    KVM over IP.

    Lantronix-Spider-screenshot

    Got my Lantronix KVM-over-IP device working, at last. Finally a way to replace my ancient and unreliable KVM over IP PCI card I've been using in an ancient computer, for about 16 years. Better ESXi how-to videos to come! Read more herehere, and here. I'll need to write up some of the parts I bought at various 3rd party sites, going with risky eBay used version to save quite a bit of money too. Got it working about an hour ago. Nice! I realize I may have built-in similar abilities in future motherboards with IPMI, but I like that I can use this on anything, even laptops. Great for producing videos that show reboots, such as: Build your own VMware vSphere ESXi 5.5 Datacenter, starting with one PC or LSI 9265-8i 5.5 has arrived, time to update MegaRAID, firmware, and SMIS right from your VM!

    LSI RAID.

    Got my LSI 9265-8i RAID adapter working great for ESXi 5.5, with MegaRAID UI now working in a Windows 8.1 VM as well. Just took some very careful tweaks to the hosts file (of ESXi and Windows), read more here.

    VMFS corruption? Nope! VMware Support saves my bacon, manually.

    Thought I had corrupted a VMFS filesystem holding some massive Windows NTFS drives in vZilla daily backups VM. Way bigger than 2TB actually, thanks to ESXi 5.5 ability to blow past that barrier. You know the sinking feeling, when you can't power up a VM, getting a variety pack of reasons and warnings. I managed to finagle an SR# with VMware Support, and tada, in 3 hours, remote support came in, and worked hard. Seeing things like Corrupt heartbeat detected doesn't instill confidence. But persistence, and 3 hours later, and it was fixed. Entirely. No data lost. At all. No need for lengthly restores. Even got some video and screen grabs as VMware worked their magic. Learned some good stuff. And learned how crummy eSATA can be, if bumped even slightly during operation. It was a friendly little reminder of another eSATA issue last year. See also my related post over on VMware Community Forums: ESXi 5.5 issue with powering on a VM:  "An error was received from the ESX host while powering on VM" "22 (Invalid argument)" "Module DiskEarly power on failed."

    Corrupt-heartbeat-detected

    Another great support experience.

    Thought my 16 year old Genie Medallion Screwdrive garage door opener was on the fritz. Nope! This model has lifetime coverage. Amazing customer service, with a pleasant support experience, wow! No need to replace, always better to repair. Phew!

    Thank you Genie, for hiring and keeping excellent technical support representatives. It will make all the difference, when the time comes for me to choose the replacement.

    Unplanned iPhone 5 Replacement, yet another amazing customer support experience.

    Died the evening before a big customer site meeting the next morning. Quickly got the last Genius Bar appointment of the day. Arrived at the closest Apple store, just before closing. The Genius bar got me going again, with amazing grace and speed. Very good customer service. Any phone replacement these days is made much more painful due to dual factor authentications, with numerous issues I fought for days to come. But I took the opportunity to take some notes, and learned a bunch about what to do next time I turn in a phone, especially if it's in working condition next time around.

    VOIP phone lines.

    SB6141-narrow-picture

    Had some nasty noise on one of my 2 Cox lines. Instead of nailing down which of many duplex RJ11 outlets in my home was to blame, I recalled that my Ooma Telo didn't exhibit that issue. I want to ditch my somewhat flakey DOCSIS 3.0 Cisco DPQ3212 anyway, and its 2 RJ11 voice ports. Why? Unreliable these past few months, needing a reboot every week or two. So dumping these 2 VOIP ports lets me go back to a pure Motorola/Arris modem with no voice ports where I can view the log files, handy for outages like last week, where my Cisco modem rebooted several times in one day. Even Cox says they don't have access to the Cisco's logs either. Not good. Sot it's back to the nice little Motorola/Arris SB6141. So here's the plan, half-way implemented:

    • Use homeservershow.com's Ooma affiliate link to re-sign up for Ooma, on that rarely used unlisted home line 1, see also Why I gave up on Ooma VOIP, and went back to cable phone service. I'll live with the lag, on this occasional-use line.
    • Have Cox business telephone service installed for my home's line 2 under my LLC, with a separate eMTA device Cox provides (happening this Thursday). We'll see how it works out. But for now, line 1 noise is now behind me. The promise of simultaneous ring, from a low latency cable company carrier grade VOIP, is ahead of me. Of course, I'll be testing it, and have 30 days to return it all, if need be.

    Automated, offsite virtual machine backups (ESXi and Hyper-V).

    Across town, my parent's house has good internet. Perfect for building out my personal cloud. So the next project is to test out a 3rd party ESXi product for offsite backups. With dedupe, and WAN optimization. The one I'm testing didn't initially like the lack of reverse DNS that's typical in a home lab. I got around that using the same host file tweaks to fix MegaRAID in an ESXi 5.5 VM. Stay tuned!