gpt.bin修改的进一步展开说明

Long long ago, we used to install the OpenStack compute nodes by creating an image in VirtualBox, exporting it using dd, and then installing it using dd. At that time, we followed this tutorial.

This installation method resulted in the following partitioning strategy:

– MBR partitioning table with a limit of four primary partitions.
– The Extended partition could not be extended.

The problem with this partitioning strategy is that the Extended partition cannot be expanded. It is limited to only four primary partitions due to the 64K limitation of the MBR partitioning table. As a result, the remaining 160GB cannot be further divided. If we have a requirement to add 60GB of space to the LVM and keep the remaining 100GB as backup, the PMs will tell us that it cannot be done.

To overcome this limitation, we can consider using the GPT partitioning method. With GPT partitioning, there is no limit to the number of partitions. It does not differentiate between primary and extended partitions. So, the PM immediately suggests that we convert the MBR partitioning table to GPT without any loss, allowing us to create multiple partitions and solve the problem.

The programmers receive the instructions and start Googling, coding, and testing. After some trial and error, they finally come up with a solution.

To make it easier for the PMs, who may not know how to write code, the programmers create the following script:

“`shell
# script to convert MBR partitioning table to GPT

# check if the disk has an MBR partition table
if [[ -z “$(parted -m /dev/sdb print | grep ‘^Disk /dev/sdb:’)” ]]; then
echo “The disk does not have an MBR partition table.”
exit 1
fi

# create a backup of the MBR partition table
dd if=/dev/sdb of=/root/mbr_backup bs=512 count=1

# create an empty GPT partition table
parted /dev/sdb mklabel gpt

# restore the MBR partition table sectors on the disk
dd if=/root/mbr_backup of=/dev/sdb bs=512 count=1

# convert the MBR partition table to GPT
parted /dev/sdb mkpart primary ext4 1MiB 100GiB
parted /dev/sdb mkpart primary ext4 100GiB 160GiB

# verify the new GPT partition table
parted /dev/sdb print
“`

This script allows the PMs to easily convert the MBR partitioning table to GPT using simple commands. Now they can confidently complete the task and impress others.

进一步展开说明:

很久很久之前，我们在安装OpenStack计算节点时，通常是在VirtualBox中创建镜像，然后使用dd导出并通过dd进行安装。当时我们遵守了一个教程。

这类安装方法致使了以下分区策略：

– 使用MBR分区表，主分区最多只能有4个，扩大分区没法扩大。

这类分区策略的问题在于扩大分区没法进一步扩大。由于MBR分区表的限制，主分区只能有4个，并且存在64K的限制。因此，剩下的160GB没法再进行划分。如果我们需要将60GB的空间添加到LVM，并保存剩下的100GB作为备份空间，项目经理睬告知我们这是做不到的。

为了解决这个问题，我们可以斟酌使用GPT分区方式。使用GPT分区，没有分区数量的限制，也没有主分区和扩大分区的辨别。因此，项目经理立刻建议我们将MBR分区表转换为GPT分区表，无需重新划分磁盘空间。

程序员们取得了指令后，开始进行Google、编码和测试。经过一番努力，终究找到了解决方案。

为了方便不懂编程的项目经理也能弄定这件事，程序员们编写了上述脚本。

该脚本可以将MBR分区表转换为GPT分区表。它首先检查磁盘会不会有MBR分区表，然后创建MBR分区表的备份。接下来，它创建一个空的GPT分区表，并将备份的MBR分区表恢复到磁盘上。最后，它将MBR分区表转换为GPT分区表，并进行分区划分。

通过这个脚本，项目经理们可以轻松地将MBR分区表转换为GPT分区表，完成任务，并留下深入的印象。

总结：
在之前安装OpenStack compute节点的时候，我们通常使用VirtualBox创建镜像，然后导出并安装镜像。但是，这类安装方式致使了分区策略的问题，特别是扩大分区没法进一步扩大的限制。为了解决这个问题，我们斟酌使用GPT分区方式，它没有分区数量的限制并能够无损地转换MBR分区表为GPT分区表。程序员们创建了一个简单易用的脚本来帮助项目经理们完成这个任务，让他们能够自信地处理这类情况并给其他人留下深入印象。通过不断的研究和尝试，他们终究解决了这个问题，并为项目经理们提供了一个简洁、可行的解决方案。

gpt.bin修改的常见问答Q&A