Yet another personal KB
Allright, this was a tricky one. Zimbra uses Postfix, but one cannot just manually edit the configuration files, as zmconfigd will be more than happy to overwrite your changes. So, after installing policyd-spf-python, using the source of course, as Zimbra does READ MORE
By default, ser2net uses the tcp accepter. This means that your input will be echo’d back to your terminal. For most use cases, that sucks. Modify it to the telnet accepter, and the issue is gone. For example, use: connection: READ MORE
So, I found out the hard way that track_instance_changes = false in nova.conf does not work. In /usr/lib/python3/dist-packages/oslo_service/periodic_task.py, add if task_name == “_sync_scheduler_instance_info”: return idle_for in the function run_periodic_tasks(), and get shit over with.
Add a system mac address to /mnt/flash/system_mac_address: echo “0000.1234.5678” > /mnt/flash/system_mac_address Add serial number info to /mnt/flash/veos_config: echo “SYSTEMMACADDRE=0000.1234.5678” >> /mnt/flash/veos_config echo “SERIAL=JPE000001” >> /mnt/flash/veos_config echo “SERIALNUMBER=JPE000001” >> /mnt/flash/veos_config
By default, at least up until CloudEOS 4.28.2F, Arista CloudEOS will not run on Xenserver as it does not recognize the environment. This is easily fixable by modifying the BIOS strings for the VM. HOWEVER…
I recently used Reportlab to generate PDFs. Unfortunately, they were deemed corrupt by my Mac’s preview and Adobe Reader. Interestingly, they would render properly on Chrome’s PDF reader, but when I tried to print, the page was empty. Long story READ MORE
wget https://cloud-images.ubuntu.com/bionic/current/bionic-server-cloudimg-amd64.img sudo modprobe nbd max_part=1 sudo qemu-nbd –connect=/dev/nbd0 bionic-server-cloudimg-amd64.img sudo mount /dev/nbd0p1 /mnt/nbd/ -o noatime pushd /mnt/nbd sudo mount -t proc proc proc sudo mount -t sysfs sysfs sys sudo mount -t devtmpfs devtmpfs dev sudo mkdir -p dev/pts READ MORE
If your image has UFS filesystems, qemu-nbd on Linux is useless. Here is what you do (example with a vsrx image): qemu-img convert -f qcow2 -O raw junos-vsrx3-x86-64-18.4R1.8.qcow2 junos-vsrx3-x86-64-18.4R1.8.img mdconfig -a -t vnode -f junos-vsrx3-x86-64-18.4R1.8.img -u 0 mount /dev/md0p2 /mnt READ MORE
First, convert the image to raw: qemu-img convert -p -f qcow2 -O raw jinstall-vqfx-10-f-17.4R1.16.img jinstall-vqfx-10-f-17.4R1.16.img.raw Next, mount it on FreeBSD: mdconfig -a -t vnode -u 0 -f jinstall-vqfx-10-f-17.4R1.16.img.raw mount /dev/md0s1a /mnt Next, go to /mnt/etc/config and edit vqxf-10000-factory.conf root@dora:/mnt/etc # READ MORE
Like this: modprobe nbd cd /var/lib/nova/instances/bed8d7bc-cde2-4bdf-9a6b-ccbd9df70912 qemu-nbd -c /dev/nbd0 disk partprobe /dev/nbd0 mount /dev/nbd0p2 /mnt cd /mnt/etc/init.d/ mv rc.platform.manual rc.platform cd /var/lib/nova/instances/bed8d7bc-cde2-4bdf-9a6b-ccbd9df70912 umount /mnt qemu-nbd -d /dev/nbd0 virsh start instance-00000325 <serial type=’pty’> <target port=’0’/> </serial> <console type=’pty’> <target type=’serial’ READ MORE