Is it safe to just remove files from /var/cache on a running system, or is there a correct procedure for doing that? Mine has hit over 3 gigs, making it one of the larger directories in /, which is running low on space. I've hit all the low-hanging fruit I can find and now I come to things like /var/cache, and I don't know what to do about such. Thanks in advance! Fred
The FHS specification says that applications using /var/cache should expect those data to disappear anytime, so under a strict interpretation it should be safe to remove everything. Practically, though, I don't know that I'd trust every application to adhere to that, so it might be worth looking at what applications are the biggest users of the directory and making targeted deletions or even application-specific tools to remove the data (i.e. "yum clean all"). On Thu, Dec 30, 2021 at 06:20:15PM -0500, Fred wrote:> Is it safe to just remove files from /var/cache on a running system, or is > there a correct procedure for doing that? > > Mine has hit over 3 gigs, making it one of the larger directories in /, > which is running low on space. I've hit all the low-hanging fruit I can > find and now I come to things like /var/cache, and I don't know what to do > about such. > > Thanks in advance!-- -- Skylar Thompson (skylar2 at u.washington.edu) -- Genome Sciences Department (UW Medicine), System Administrator -- Foege Building S046, (206)-685-7354 -- Pronouns: He/Him/His
--On Thursday, December 30, 2021 6:20 PM -0500 Fred <fred.fredex at gmail.com> wrote:> Mine has hit over 3 gigs, making it one of the larger directories in /, > which is running low on space. I've hit all the low-hanging fruit I can > find and now I come to things like /var/cache, and I don't know what to do > about such.Have you run KDirStat to find the disk pigs? I regularly use its Windows derivative, WinDirStat. <http://kdirstat.sourceforge.net/> <https://windirstat.net/> I wish there was a web-based version for headless servers without X libraries.
On Thu, 30 Dec 2021 at 18:21, Fred <fred.fredex at gmail.com> wrote:> Is it safe to just remove files from /var/cache on a running system, or is > there a correct procedure for doing that? > > Mine has hit over 3 gigs, making it one of the larger directories in /, > which is running low on space. I've hit all the low-hanging fruit I can > find and now I come to things like /var/cache, and I don't know what to do > about such. > >The first step is to find out what is using it. It is probably dnf but could be other utilities which are trying and failing to do something. I start off with ``` $ sudo -i # cd /var/cache # du -sch | sort -h 0 ./PackageKit 0 ./app-info 0 ./bpf 0 ./fwupd 0 ./httpd 0 ./krb5rcache 0 ./libX11 0 ./libvirt 0 ./private 0 ./realmd 36K ./ldconfig 1.7M ./man 29M ./dnf 31M total while on a different system: 4.0K ./abrt-di 4.0K ./bpf 4.0K ./foomatic 4.0K ./krb5rcache 4.0K ./private 4.0K ./realmd 8.0K ./httpd 8.0K ./libX11 8.0K ./powertop 96K ./ldconfig 300K ./ibus 520K ./libvirt 3.5M ./man 4.2M ./fwupd 38M ./app-info 59M ./cups 213M ./PackageKit 332M ./dnf 2.1G ./mock 2.7G total ``` As others have noted, dnf is probably the most used tool here, but it could be mock or some other utility (I had cups because I misconfigured something once) dnf is a tricky tool because sometimes a command will create a 'not-so-temporary' cached tree which can't be cleaned because `dnf clean all` doesn't know it. What I do is a `dnf clean all` and then go into /var/cache/dnf and see what else might be still there. In my case I found a large trove of packages from when I had enabled testing at one point and then turned it off before doing a clean. I normally just delete all the directories and do a `dnf update` to see if it reports errors. Hope this helps. Thanks in advance!> > Fred > _______________________________________________ > CentOS mailing list > CentOS at centos.org > https://lists.centos.org/mailman/listinfo/centos >-- Stephen J Smoogen. Let us be kind to one another, for most of us are fighting a hard battle. -- Ian MacClaren
well, I removed all the files in the tree under /var/cache/yum/x86_64/7 but left all the directories empty. that got rid of a couple gigs of stuff. among the remains, the only other big one remaining is: 2.3G abrt-di which I won't mess with for now. I've got 4.2G free, now, so that should run me for quite a while. Thanks to all of you for your tips! Fred On Fri, Dec 31, 2021 at 10:38 AM Stephen John Smoogen <smooge at gmail.com> wrote:> On Thu, 30 Dec 2021 at 18:21, Fred <fred.fredex at gmail.com> wrote: > > > Is it safe to just remove files from /var/cache on a running system, or > is > > there a correct procedure for doing that? > > > > Mine has hit over 3 gigs, making it one of the larger directories in /, > > which is running low on space. I've hit all the low-hanging fruit I can > > find and now I come to things like /var/cache, and I don't know what to > do > > about such. > > > > > The first step is to find out what is using it. It is probably dnf but > could be other utilities which are trying and failing to do something. I > start off with > > ``` > $ sudo -i > # cd /var/cache > # du -sch | sort -h > 0 ./PackageKit > 0 ./app-info > 0 ./bpf > 0 ./fwupd > 0 ./httpd > 0 ./krb5rcache > 0 ./libX11 > 0 ./libvirt > 0 ./private > 0 ./realmd > 36K ./ldconfig > 1.7M ./man > 29M ./dnf > 31M total > > while on a different system: > 4.0K ./abrt-di > 4.0K ./bpf > 4.0K ./foomatic > 4.0K ./krb5rcache > 4.0K ./private > 4.0K ./realmd > 8.0K ./httpd > 8.0K ./libX11 > 8.0K ./powertop > 96K ./ldconfig > 300K ./ibus > 520K ./libvirt > 3.5M ./man > 4.2M ./fwupd > 38M ./app-info > 59M ./cups > 213M ./PackageKit > 332M ./dnf > 2.1G ./mock > 2.7G total > > ``` > > As others have noted, dnf is probably the most used tool here, but it could > be mock or some other utility (I had cups because I misconfigured something > once) > > dnf is a tricky tool because sometimes a command will create a > 'not-so-temporary' cached tree which can't be cleaned because `dnf clean > all` doesn't know it. What I do is a `dnf clean all` and then go into > /var/cache/dnf and see what else might be still there. In my case I found a > large trove of packages from when I had enabled testing at one point and > then turned it off before doing a clean. I normally just delete all the > directories and do a `dnf update` to see if it reports errors. > > Hope this helps. > > > > Thanks in advance! > > > > Fred > > _______________________________________________ > > CentOS mailing list > > CentOS at centos.org > > https://lists.centos.org/mailman/listinfo/centos > > > > > -- > Stephen J Smoogen. > Let us be kind to one another, for most of us are fighting a hard battle. > -- Ian MacClaren > _______________________________________________ > CentOS mailing list > CentOS at centos.org > https://lists.centos.org/mailman/listinfo/centos >