a day full of failure, with a final success!

Storage 

Finally received the SAS cable for my H700 card, slotted in the H700 card and attempted to boot up my drives which were on PERC 6i previously. Failed to import foreign config....... cant find much relevant info on Google. gave up, rolled back to EPERC 6i.

I think this is a lesson learnt that it really is not a good idea to put freeNAS's ZFS storage on a RAID card, i have read many resource asking people not to do so due to various reasons but I chose to ignore it for convenience. 

Alright. So I am thinking if i should get a H200 card flashed in IT mode to host the storage drive for freeNAS. But at the same time I felt I spent too much time fucking around with this bunch of system stuff which really is not my focus professionally. But at the same time i cant take the fact that I have a shitty design. DIEMMA.    anyway found this video - https://www.youtube.com/watch?v=YYP62rPoXwI  splitting backplane on Dell R710, drive 1 - 4 on HW RAID, drive 5 - 6 on HBA for freeNAS, seems like what I just need....hmmmm..

GPU / PCI passthrough.

Finally received my x8 PCIe GT710 graphics card, it was not a common product since majority of the graphics card nowadays are more then x8. however the Dell R710 can only support x8 max. and I dont wanna mess around with putting a x16 car don x8 slot by cutting here and there or adding riser card, I wanna keep it standard and neat, so I got this Dell branded card from taobao which is pretty cheap.

Slotted it in and spent the whole day trying to get PCI pass-through working, but had trouble turning on iommu, troubleshooted for the whole day, doubted myself, doubted the machine ( doubt that it doesnt support vt-d but its nonsense since i have eve-ng and free-nas running), finally turns out it is this fucking stupid bug https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1569567 from 2016. I cant believe, ad I dont know how am I still affected in 2020. 

Basically any parameter you added on boot/grub/default is IGNORED by update-grub, the resultant grub.cfg does not contain anything you added. The workaround is grub-mkconfig -o /boot/grub/grub.cfg

I am really furious that such basic config is still troubled by such low level bug for so many years. Anyway.

🧡

root@eve-ng:/# dmesg | grep IOMMU

[    0.503059] DMAR: IOMMU enabled

[    0.503121] DMAR: IOMMU enabled

🧡🧡

root@eve-ng:/# find /sys/kernel/iommu_groups/ -type l | wc -l

76

🧡🧡🧡

root@eve-ng:/# dmesg | grep -i vfio

[    0.000000] Command line: BOOT_IMAGE=/vmlinuz-4.20.17-eve-ng-ukms+ root=/dev/mapper/eve--ng--vg-root ro net.ifnames=0 intel_iommu=on vfio_iommu_type1.allow_unsafe_interrupts=1 quiet intel_iommu=on vfio_iommu_type1.allow_unsafe_interrupts=1

[    0.502975] Kernel command line: BOOT_IMAGE=/vmlinuz-4.20.17-eve-ng-ukms+ root=/dev/mapper/eve--ng--vg-root ro net.ifnames=0 intel_iommu=on vfio_iommu_type1.allow_unsafe_interrupts=1 quiet intel_iommu=on vfio_iommu_type1.allow_unsafe_interrupts=1

[   14.319954] VFIO - User Level meta-driver version: 0.3

[   14.635412] vfio_pci: add [10de:128b[ffff:ffff]] class 0x000000/00000000

[   14.635428] vfio_pci: add [10de:0e0f[ffff:ffff]] class 0x000000/00000000

[ 2161.807383] vfio-pci 0000:06:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=none

[ 2169.203371] vfio-pci 0000:06:00.0: enabling device (0000 -> 0003)

🧡🧡🧡🤍

one last obstacle, error 43, for tomorrow.

L3 port channel doesnt work on iol L2 15.2d

SW1#sho ver | i bin
System image file is "unix:/opt/unetlab/addons/iol/bin/i86bi-linux-l2-adventerprisek9-15.2d"
SW1#sho etherchannel summ | i Po1
1      Po1(RU)          -        Et1/2(P)    Et1/3(P) 
SW1#sho ip int bri | i 12.0.0
Port-channel1          12.0.0.1        YES manual up                    up   
SW1#sho arp | i 12.0.0.
Internet  12.0.0.1                -   aabb.cc80.b000  ARPA   Port-channel1
SW1#ping 12.0.0.2
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 12.0.0.2, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)

SW2#sho ver | i bin
System image file is "unix:/opt/unetlab/addons/iol/bin/i86bi-linux-l2-adventerprisek9-15.2d"
SW2#sho etherchannel summ | i Po1
1      Po1(RU)          -        Et1/2(P)    Et1/3(P) 
SW2#sho ip int bri | i 12.0.0
Port-channel1          12.0.0.2        YES manual up                    up   
SW2#sho arp | i 12.0.0.
Internet  12.0.0.1                4   aabb.cc80.b000  ARPA   Port-channel1
Internet  12.0.0.2                -   aabb.cc80.c000  ARPA   Port-channel1
SW2#ping 12.0.0.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 12.0.0.1, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)




googled a bit, not much relevant result, only saw some mention about IOL L2 being notoriously buggy. was annoyed to a point that i went on to search for physical switch on taobao again. after a while went back to consult the official requirement: https://learningnetwork.cisco.com/s/article/ccie-enterprise-infrastructure-equipment-and-software-list
FUCK, it is actually talking about IOSv all the while, not IOL, so i've been using the wrong image, thats why it gave me a hard time for anything to do with layer3, SVI. wasted so much time doubting my own config.
FUCK.

Virtual machines

• Cisco CSR 1000v Series Cloud Services Routers with Cisco IOS XE SD-WAN Release 16.12
• Cisco IOSv with Cisco IOS Software Release 15.8
• Cisco IOSv-L2 with Cisco IOS Software Release 15.2
• Cisco SD-WAN (vManage, vBond, vSmart, vEdge) Software Release 18.4
• Cisco DNA Center Release 1.3.1

~

Switch#sho ver | i ios         
Cisco IOS Software, vios_l2 Software (vios_l2-ADVENTERPRISEK9-M), Experimental Version 15.2(20170321:233949) [mmen 101]
System image file is "flash0:/vios_l2-adventerprisek9-m"
Switch#sho etherchannel s | i Po1
1      Po1(RU)          -        Gi0/0(P)    Gi0/1(P)   
Switch#sho ip int bri | i 1.0
GigabitEthernet1/0     unassigned      YES unset  up                    up     
Port-channel1          1.0.0.1         YES manual up                    up     
Switch#sho arp | i 1.0.0.   
Internet  1.0.0.1                 -   5000.001e.8000  ARPA   Port-channel1
Internet  1.0.0.2                 1   5000.001f.8000  ARPA   Port-channel1
Switch#ping 1.0.0.2
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 1.0.0.2, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 2/2/2 ms


Switch#sho ver | i vios
Cisco IOS Software, vios_l2 Software (vios_l2-ADVENTERPRISEK9-M), Experimental Version 15.2(20170321:233949) [mmen 101]
System image file is "flash0:/vios_l2-adventerprisek9-m"
Switch#sho etherchannel s | i Po1
1      Po1(RU)          -        Gi0/0(P)    Gi0/1(P)   
Switch#sho ip int bri | i 1.0
GigabitEthernet1/0     unassigned      YES unset  up                    up     
Port-channel1          1.0.0.2         YES manual up                    up     
Switch#sho arp | i 1.0.0
Internet  1.0.0.1                 2   5000.001e.8000  ARPA   Port-channel1
Internet  1.0.0.2                 -   5000.001f.8000  ARPA   Port-channel1
Switch#ping 1.0.0.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 1.0.0.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/1/3 ms

STP bug on IOL l2 image

Past 2 days was spent mainly to get familiar with secureCRT & Anki.

Today:

- converted my hand written note on some INE SD-WAN & L2 videos into Anki card

- Discovered a STP bug with i86bi-linux-l2-adventerprisek9-15.6.0.9S.bin, STP does not converge, 2 switches connected with 2 trunk link => both elect themselves as root and eventually crash 👀😕. downgraded to i86bi-linux-l2-adventerprisek9-15.2d.bin, all good. should have paid attention to the suggested image version on https://learningnetwork.cisco.com/s/article/ccie-enterprise-infrastructure-equipment-and-software-list. DONT BE ADVENTUROUS !!

Restarting

So I was Googling about CCIE exam prep stuff and was ecstatic to find https://nickccieprogress.blogspot.com/ and couldn't stop myself from leaving a note of thanks(which I usually only do if i am super super thankful). Then i was a little surprised to find that i have a "blogger" account already existed. a few more clicks and found myself back to this blog, apparently, back in 2011 when owning a macbook was a new hype and I created this blog to share knowledge / experience. 

The idea of studyingCCIE has been floating around my head for about a year but for most time i denied the desire/need of doing so, plus various other reasons.

I have also started the idea of building a homelab since July 2019, that was mainly network / hardware / casual / for fun focused; I bought an Aruba 650 controller with 2 AP95, palo alto PA-200. and was trying to hook them up with virtual topology inside GNS3, had lots of problem with GNS3, to a certain point I started looking at eve-ng, and was surprised how easy it was, how well documented the howto's are. That gave me confident that I can manage a lab environment for learning, the only problem left was PC resource(CPU cores/ ram) and reliability (my new job requires me to travel all the time, can I establish an environment thats super reliable and allow remote access any time?) so i started looking at home server options, during the process i decided to make a server that can reliably run labs & replace my old Synology NAS

Then I did some research and bought a 2nd hand Dell R710, spent some time learning and researching on the right hypervisors for myself. tested hyper-v, XCP, and finally decided to run eve-ng baremetal as thats the easiest and provide "almost guaranteed" compatibility. During the process Im surpised to be able to get instant help from eve-ng community helpdesk, thats as good as, or even better, than internal IT helpdesk of companies i have worked for.

After a few weeks of struggle i can say that my environment is stable enough, and during the process the covid19 / wuhan virus pandemic started, was thinking how to best utilize my time. My new job gave me a very good motivation / momentum in learning new stuff and being more technical, together with a reliable server running in front of me, I think it is the perfect time to start my study.

During this whole process, I struggled a lot with the technical bits, and have always been thinking about documenting it somewhere. Rediscovering this blog I had back in 2011, is simply like "Hey im here, Im still alive! use me!" So here we go.



Done so far:
- Installed UPS for physical lab setup
- eve-ng environment fully ready
- freeNAS fully setup, OS is running on top of KVM, storage is 2 * 4TB drive directly attached to freeNAS to be managed as ZFS, data migration completed, SMB / NFS working reliably. offsite backup to AWS S3 in place
- remote access fully ready and tested on all my laptops & Android
- INE subscription in, workbook ready
- Safari book subscription in
- CCIE v5 required topology preliminary ready on eve-ng with IOL L2/L3 devices, INE workbook lab initial config ready
- Anki installed on PC & Android with deck from neckercube and Nick

To do:
- read that some required features may not be available on IOL. I do have CSR1000V 16.x in working condition however the eve-ng lab file from Nick is using CSR1000V 3.x so im unable to fire that up, not sure how to modify it efficiently so I will live with IOL setup for now, believe it will take a while for my learning to progress to a stage that will be hindered by the IOL limitation.
- upgrade Dell R710 from PERC to H700 to support >2TB drives, already bought the card but I bought the wrong SAS cable, not very urgent as i can still live with 2TB drive.
- add another HDD to run RAID 1 for the main disk on the Dell server for fault tolerant. Actually quite critical but I'm inclined to do this after upgrade to H700, as I dont risk migrating an array from one controller to another, I have no experience on that.

Adobe Photoshop Alert "Delete the Adobe Photoshop Settings file?" | Tools, functions behave strangely [Temp. resolution]

Problem raised: Mid-Nov 2010

Problem solved: Mid-Dec 2011
Solution type: unofficial solution

Description: After a full restore from my time machine backup, my photoshop CS3 doesn't behave normally, the warning, "Delete the Adobe Photoshop Settings file?" appears when I open Photoshop, and the tools behave as if the Shift or Option keys are pressed. reinstalling doesn't work, and upgrading to CS5 also cant solve the problem. Photoshop basically cant work as all the tools behave abnormally.

Causes: There appears to be a conflict between Mac OS 10.6.5, Photoshop, and some third-party utilities that alter keyboard shortcuts. These third-party utilities include Typinator, TextExpander, moveAddict, Things, xGestures, and so on. 

Solution: There is a temporary resolution by uninstalling moveAddict or any other third-party utilities stated above

Credit:
Adobe KB

Plantronics Voyoger Pro not working on mac [solved]

Problem raised: Mid-Oct 2010
Problem solved: 2 Jan 2011
Solution type: unofficial solution

Description: I have bought a Plantronics Voyoger Pro for use with skype on mac at around august 2010, it worked very well until around October when it suddenly unable to input/output audio signal. It is still able to pair with the bluetooth module on mac, but when i try to play any sound from my mac, i am not able to hear any sound from the headset, and the beach ball appears on the mac, with some applications becoming unresponsive.

Causes: the Mac OSX 10.6.5 update includes an update for the bluetooth audio agent, with makes some bluetooth headset unusable with mac

Solution: rollback the bluetooth audio agent to an earlier version

1. locate the current bluetooth audio agent(BluetoothAudioAgent.app) in "/System/Library/CoreServices/", My current version is 2.3.8.

2. Get a previous version of the BluetoothAudioAgent.app and replace the current file on your mac, my previous version was 2.3.3. You can get the previous version of the file from either one of the following ways.

1. http://rapidshare.com/files/440364260/BluetoothAudioAgent.app.zip
2. Your time machine backup
3. MacOSX Installation disk (Get Pacifist to browse the package content of the disk, and find the BluetoothAudioAgent.app from the package contents)

3. Try to delete the bluetooth device and pair it again, it should work like what it did in the past!

Credit:
caseno7358 @ apple discussion forum

Hello World

This is a blog for sharing my experience on solving some tricky problems on Mac OS X. I am from Hong Kong, I am not really an expert on Mac, just a normal user like most of you are.

There is always times when you encounter some problems (on Mac) that have the following characteristics:
1. Not a serious problem - you can still use your Mac, but with some glitches
2. It become serious at times - when you come across a certain functionality, the problem makes you annoyed
3. You can search the question on Google, but not the answer - Mac is nice, you like it, and so do I. But it is undeniable that Mac user is still a small community, and it takes time for the community to discover, analyze, and solve the problem. It is even harder to make the problem noticeable by Apple, and see its action.

My experience is that many users are asking the same question on the Internet, even if there is a workable solution somewhere, it takes a long time to propagate through the whole community. With this blog, I would like to share some of my result in searching of solution for problems related to Mac, the solution may not be original, it mainly serves the purpose of reducing your effort on solving minor problems on Mac, and foster knowledge sharing among Mac users. I will try my best to credit the source.

I welcome your comment and it is even better if you can share your experience too.