Bug 468291
Summary: | tmpwatch shouldn't run when an nfs directory is mounted on /tmp | ||||||
---|---|---|---|---|---|---|---|
Product: | Red Hat Enterprise Linux 5 | Reporter: | Jonathan Schatz <jon> | ||||
Component: | tmpwatch | Assignee: | Miloslav Trmač <mitr> | ||||
Status: | CLOSED WONTFIX | QA Contact: | BaseOS QE <qe-baseos-auto> | ||||
Severity: | medium | Docs Contact: | |||||
Priority: | medium | ||||||
Version: | 5.4 | ||||||
Target Milestone: | rc | ||||||
Target Release: | --- | ||||||
Hardware: | All | ||||||
OS: | Linux | ||||||
Whiteboard: | |||||||
Fixed In Version: | Doc Type: | Bug Fix | |||||
Doc Text: | Story Points: | --- | |||||
Clone Of: | Environment: | ||||||
Last Closed: | 2008-10-24 01:33:43 UTC | Type: | --- | ||||
Regression: | --- | Mount Type: | --- | ||||
Documentation: | --- | CRM: | |||||
Verified Versions: | Category: | --- | |||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
Cloudforms Team: | --- | Target Upstream Version: | |||||
Embargoed: | |||||||
Attachments: |
|
Description
Jonathan Schatz
2008-10-23 21:59:00 UTC
Created attachment 321359 [details]
Patch to keep tmpwatch from running on nfs mounts
Thanks for your report. I'm afraid I don't think this patch is generally useful to tmpwatch users. Some systems have no local storage and use NFS or other network file systems for all mounts. Although mounting a NFS storage over /tmp is unusual, tmpwatch must be useful in other cases as well, e.g. to automatically remove old data from a shared file system that contains daily software builds. Thus, the default behavior should definitely not exclude NFS mounts. That leaves the possibility of adding a --nonfs option, to let users modify /etc/cron.daily/tmpwatch and disable this behavior if it is likely to happen on their systems. Such change does not belong in tmpwatch either: First, /etc/cron.daily/tmpwatch can already be modified not to run on NFS mount points even without the --nonfs option by checking /etc/fstab in the shell script. Second, mounting a NFS directory over /tmp is only one of many ways to break an UNIX system (e.g. bind mounting /var - instead of /var/tmp - over /tmp, or deleting /dev/null or /bin/sh). I don't think it makes sense to specifically handle one or two such cases in tmpwatch, tmpwatch can never provide any reasonable assurance that it will not make things worse on an incorrectly used system. |