System disk partitioning
Ste
125 Posts
January 21, 2022, 8:43 amQuote from Ste on January 21, 2022, 8:43 amHello, yesterday the root partition of one petasan node filled up to 100%, because the /var/log/samba/log.ctdb file grew up to 5 GB in three months. Well, I easily solved clearing that file, but I was a bit surprised of this, because I dedicated a 128 GB SSD to system disk. Later I realized that most of the disk is mostly allocated with almost empty partitions:
Filesystem Size Used Avail Use% Mounted on
udev 32G 0 32G 0% /dev
tmpfs 6.3G 2.6M 6.3G 1% /run
/dev/sdk3 15G 9.8G 5.0G 67% /
tmpfs 32G 84K 32G 1% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 32G 0 32G 0% /sys/fs/cgroup
/dev/sdk4 30G 57M 30G 1% /var/lib/ceph <-----
/dev/sdk5 66G 54M 66G 1% /opt/petasan/config <-----
I''m wandering if there's a reason to give all that space to /opt/petasan/config and /var/lib/ceph, leaving the logs directory on the root filesystem, and if it's possible to give the user the opportunity to customize the system disk partitioning at install time, expecially for large disks.
Thanks and bye. Ste.
Hello, yesterday the root partition of one petasan node filled up to 100%, because the /var/log/samba/log.ctdb file grew up to 5 GB in three months. Well, I easily solved clearing that file, but I was a bit surprised of this, because I dedicated a 128 GB SSD to system disk. Later I realized that most of the disk is mostly allocated with almost empty partitions:
Filesystem Size Used Avail Use% Mounted on
udev 32G 0 32G 0% /dev
tmpfs 6.3G 2.6M 6.3G 1% /run
/dev/sdk3 15G 9.8G 5.0G 67% /
tmpfs 32G 84K 32G 1% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 32G 0 32G 0% /sys/fs/cgroup
/dev/sdk4 30G 57M 30G 1% /var/lib/ceph <-----
/dev/sdk5 66G 54M 66G 1% /opt/petasan/config <-----
I''m wandering if there's a reason to give all that space to /opt/petasan/config and /var/lib/ceph, leaving the logs directory on the root filesystem, and if it's possible to give the user the opportunity to customize the system disk partitioning at install time, expecially for large disks.
Thanks and bye. Ste.
admin
2,930 Posts
January 22, 2022, 10:51 amQuote from admin on January 22, 2022, 10:51 amThe ceph directory contains the mon database, it can grow in case of large cluster recovery operations. The config dir contains shared PetaSAN data, currently it saves statistics charts info for 1 year but we may add other data in future. We do have log limits/rotation of services we use, but maybe ctdb slipped from us, i will double check this and make sure we add it. You can also symlink the /var/log to one the large partitions, but i think it is best for us to put log limit on ctdb.
The ceph directory contains the mon database, it can grow in case of large cluster recovery operations. The config dir contains shared PetaSAN data, currently it saves statistics charts info for 1 year but we may add other data in future. We do have log limits/rotation of services we use, but maybe ctdb slipped from us, i will double check this and make sure we add it. You can also symlink the /var/log to one the large partitions, but i think it is best for us to put log limit on ctdb.
Ste
125 Posts
January 25, 2022, 9:32 amQuote from Ste on January 25, 2022, 9:32 amOk, thanks for the explaination. If you limit the logs then I agree that your setup is the most suited.
Bye.
Ok, thanks for the explaination. If you limit the logs then I agree that your setup is the most suited.
Bye.
System disk partitioning
Ste
125 Posts
Quote from Ste on January 21, 2022, 8:43 amHello, yesterday the root partition of one petasan node filled up to 100%, because the /var/log/samba/log.ctdb file grew up to 5 GB in three months. Well, I easily solved clearing that file, but I was a bit surprised of this, because I dedicated a 128 GB SSD to system disk. Later I realized that most of the disk is mostly allocated with almost empty partitions:
Filesystem Size Used Avail Use% Mounted on
udev 32G 0 32G 0% /dev
tmpfs 6.3G 2.6M 6.3G 1% /run
/dev/sdk3 15G 9.8G 5.0G 67% /
tmpfs 32G 84K 32G 1% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 32G 0 32G 0% /sys/fs/cgroup
/dev/sdk4 30G 57M 30G 1% /var/lib/ceph <-----
/dev/sdk5 66G 54M 66G 1% /opt/petasan/config <-----I''m wandering if there's a reason to give all that space to /opt/petasan/config and /var/lib/ceph, leaving the logs directory on the root filesystem, and if it's possible to give the user the opportunity to customize the system disk partitioning at install time, expecially for large disks.
Thanks and bye. Ste.
Hello, yesterday the root partition of one petasan node filled up to 100%, because the /var/log/samba/log.ctdb file grew up to 5 GB in three months. Well, I easily solved clearing that file, but I was a bit surprised of this, because I dedicated a 128 GB SSD to system disk. Later I realized that most of the disk is mostly allocated with almost empty partitions:
Filesystem Size Used Avail Use% Mounted on
udev 32G 0 32G 0% /dev
tmpfs 6.3G 2.6M 6.3G 1% /run
/dev/sdk3 15G 9.8G 5.0G 67% /
tmpfs 32G 84K 32G 1% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 32G 0 32G 0% /sys/fs/cgroup
/dev/sdk4 30G 57M 30G 1% /var/lib/ceph <-----
/dev/sdk5 66G 54M 66G 1% /opt/petasan/config <-----
I''m wandering if there's a reason to give all that space to /opt/petasan/config and /var/lib/ceph, leaving the logs directory on the root filesystem, and if it's possible to give the user the opportunity to customize the system disk partitioning at install time, expecially for large disks.
Thanks and bye. Ste.
admin
2,930 Posts
Quote from admin on January 22, 2022, 10:51 amThe ceph directory contains the mon database, it can grow in case of large cluster recovery operations. The config dir contains shared PetaSAN data, currently it saves statistics charts info for 1 year but we may add other data in future. We do have log limits/rotation of services we use, but maybe ctdb slipped from us, i will double check this and make sure we add it. You can also symlink the /var/log to one the large partitions, but i think it is best for us to put log limit on ctdb.
The ceph directory contains the mon database, it can grow in case of large cluster recovery operations. The config dir contains shared PetaSAN data, currently it saves statistics charts info for 1 year but we may add other data in future. We do have log limits/rotation of services we use, but maybe ctdb slipped from us, i will double check this and make sure we add it. You can also symlink the /var/log to one the large partitions, but i think it is best for us to put log limit on ctdb.
Ste
125 Posts
Quote from Ste on January 25, 2022, 9:32 amOk, thanks for the explaination. If you limit the logs then I agree that your setup is the most suited.
Bye.
Ok, thanks for the explaination. If you limit the logs then I agree that your setup is the most suited.
Bye.