Сокращение дискового ввода-вывода за счет оптимизации LVE & amp; CageFS с SSD
Форум
  1. Forums
  2. General
  3. General Discussion
  1. Nick Teeple
  2. Sunday, 05 May 2013
  3.  Subscribe via email
Attempting to reduce disk IO generated by CageFS
Rate this post:
  1. 15.08.2014 07:08:51
  2. # 1
Bogdan Accepted Answer
Posts: 709
Joined: 26.06.2013
0
Votes
Undo
Hi,

The \'cagefsctl --force-update\' process really can cause high disk loads as it check or make filesystem template for all system files (for all what is written in /etc/cagefs/cagefs.mp file), if your ssd has enough capacity and power to handle something more then mysql you can really try to move /usr/share/skeleton and /var/cagefs to it, following documentation could be helpful:
http://docs.cloudlinux.com/index.html?moving_cagefs-skeleton_directo.html
http://docs.cloudlinux.com/index.html?moving__var_cagefs_directory.html

And, do you have default (unmodified) /etc/cagefs/cagefs.mp ?

About \'cldetect --cxs-installed\' it is for some intrenal use when installing cagefs.
  1. 14.08.2014 13:08:05
  2. # 2
Stéphan Schamp Accepted Answer
Posts: 4
Joined: 31.07.2014
0
Votes
Undo
Strange that nobody has answered these questions.
Could someone shed some light here?

Thanks!
  1. 05.05.2013 18:05:36
  2. # 3
Nick Teeple Accepted Answer
Posts: 1
Joined: 05.05.2013
0
Votes
Undo
Hello,

I'm working on a server with the following:
CloudLinux 5.9 (2.6.18-448.3.1.el5.lve0.8.64)

# df -h
Filesystem Size  Used Avail Use% Mounted on
/dev/sda2 9.9G  2.4G  7.0G  26% /
/dev/sda1 487M   38M  424M   9% /boot
/dev/sda8 864G  741G   80G  91% /home
/dev/sda7 2.0G  154M  1.8G   9% /tmp
/dev/sda6   20G  9.4G  9.4G  51% /usr
/dev/sda5   20G  8.8G   10G  47% /var
/dev/sdb1 1.8T  1.2T  549G  69% /backup
/dev/sdc1   74G  6.2G   64G   9% /var/lib/mysql

While watching top and iostat, I see extended periods of 100% utilization of /home and /var. These seem to correlate with the following process running and staying near the top of the top output:

/usr/sbin/cagefsctl --force-update

/dev/sdc is an SSD drive. I'm curious if anyone has any suggestions on reducing the disk IO that the above and the lvestats-server process generate with moving certain files to the SSD as one potential method of accomplishing this goal. Other methods more than welcome.

I think this may be related: cldetect --cxs-installed
I couldn't find any documentation on what this does, but I just now ran it, as cxswatch is also running on this server. Does anyone have any info on what this process does and if it can help with disk IO?

Thanks for any insight!
  • Page :
  • 1


There are no replies made for this post yet.
Be one of the first to reply to this post!
гость
Submit Your Response
Upload files or images for this discussion by clicking on the upload button below. Supports gif,jpg,png,zip,rar,pdf
• Insert • Remove Upload Files (Maximum File Size: 2 MB)
Captcha
To protect the site from bots and unauthorized scripts, we require that you enter the captcha codes below before posting your question.