New answers tagged

0

A more concise answer IMO. Your directory does not have executable permissions, which are required by cd. Fix: $ sudo chmod +x ./log Recursively: $ sudo chmod -R +x ./log +x is adding the executable attribute. You can always remove the attribute by doing -x


1

I'm not sure, but maybe your MySQL is using sparse files You can try to do your rsync with --sparse or --inplace as explained here : Rsync and sparse files


1

You could create a loopback-filesystem on each device, then group those into a raid-0. However, why not get an appropriately sized filesystem for your 80G instead of a hacked together pseudo-device?


1

You can use sysdig: sudo apt-get install sysdig; sysdig -c topfiles_bytes; There's also an interactive (top-like) terminal program that might be useful in your case. csysdig;


0

with RHEL7, one can grow file system using xfs_growfs(8), needless to say it is risky and before doing anything like that, make sure you have backup.. I asked similar question about a year ago and @Michael Hampton answer it: redhat - How to Increase the Size of an XFS File System? - Server Fault


1

Its extremely difficult to achieve what you want. I dont suggest reformatting as there is a means to salvage this. I'd drop /usr completely by copying the entire contents of it into the /usr mounted in /. Something like this should work. I haven't tested it. mkdir /tmp/reroot mount /dev/sda2 /tmp/reroot cp -a /usr/. /tmp/reroot/usr/. vi /etc/fstab # edit ...


1

You can move some content from /usr to /home with cp then soft linking the old location to the new. For example you can move the yum data folder, I already did it on some server on the past.


0

Partitions must be continuous on a disc, and as your /home and /usr aren't contiguous you can't do what you're proposing. Even if there were you would have to move the start of onw of them, that's generally very hard. It depends on the filesystem how you shrink (if you even can) or expand it, and none of output show what filesystem(s) you use.


1

Agree with Shodanshok that deadline is probably a good idea. Far from convinced that you should be using XFS here. find /volume/data/customer/ -type f -iname "*.ext" -mmin +60 -delete XFS used to be very bad with deleting files - I'm told that most of the bugs in this area have been resolved but not done any hard benchmarking to confirm this. it ...


0

You can't really resize a basic partition, since partitions are made via physical slices of the disk, so: sda 8:0 0 20G ├─sda1 8:1 0 9.8G 0 - 9.8G ├─sda2 8:2 0 4.9G 9.8G - 14,7G ├─sda3 8:3 0 1.5G 14,7G - 16,2G ├─sda4 8:4 0 1K 16,2G - 20G (LOGICAL PARTITION) └─sda5 8:5 0 1000M 16,2G(+1K) - 17,2G You ...


1

First, XFS is the right choice for this kind if scenario: with XFS is almost impossible to go out of inodes. To increase your delete performance, try the following: use the deadline I/O scheduler rather the the (default) cfq use logbsize=256k,allocsize=64k as mount options (in addition to nodiratime,noatime) To lower the deletes impact on other system ...


1

You should probably just use the XFS filesystem for this. ext4 is not well-suited to this workload because you're on an SSD and have a stupid-high file count. There's a reason Red Hat now defaults to XFS.


0

I don't have enough reputation points to add to the comments, but I would also add why not consider using ZFS on linux? ZFS is definetly superior to ext4 and it's considered production ready on linux. See http://www.phoronix.com/scan.php?page=news_item&px=MTc4NTM It's not difficult to install; see http://zfsonlinux.org/ for installation info.


2

Not the best solution, but maybe part of it, and the simplest : use suExec. With suExec, you can easily assign a different Unix user for every vhost for instance (the most common scenario). If you take care of having those user's home private (mode 0700), that's a pretty good isolation for a start. They are not chrooted, and will still share /tmp, see ...


1

Configure the approriate SELinux booleans List the related booleans # semanage boolean -l | egrep 'ftp|http' | sort Turn on booleans that may be applicable temporarily i.e. # setsebool httpd_can_connect_ftp on To turn a boolean on permanently use -P flag i.e. # setsebool -P ftp_home_dir on


1

How about "bind mount" ? # mount -o bind /var/www/vhosts/website1.com/ /home/ContractorA/website1.com/ # mount -o bind /var/www/vhosts/website2.com/ /home/ContractorA/website2.com/ # mount -o bind /var/www/vhosts/website3.com/ /home/ContractorA/website3.com/ # mount -o bind /var/www/vhosts/website1.com/ /home/ContractorB/website1.com/ # mount -o bind ...



Top 50 recent answers are included