Posts

Showing posts with the label systemd

RHEL 7 and CentOS 7 syslog Rate Limit

https://access.redhat.com/solutions/1417483 In RHEL 7 there is rate-limiting both in systemd-journald and in rsyslog's imjournal module Lower Ratelimit Interval Lower the interval for rate-limiting and increase the burst level in order to minimize the possibility of losing log messages when the threshold is reached for the specified number of messages logged within the specified interval. Rate-limiting is specific to each process, so there's usually no reason to change this. It is also inadvisable to disable this feature entirely! grep -i rate /etc/systemd/journald.conf #RateLimitInterval=30s #RateLimitBurst=1000 RateLimitInterval=10s RateLimitBurst=3000 grep -i rate /etc/rsyslog.conf #$imjournalRatelimitInterval 600 <--default $imjournalRatelimitInterval 300 $imjournalRatelimitBurst 30000 journal corruption journalctl --verify journalctl --force

Specify I/O Scheduler with udev rules

Since the advent of systemd the use of /etc/rc* startup-scripts has been discouraged and phased out. However, systemd still supports the use of certain local startup scripts for compatibility purposes. Nevertheless, to ensure full current- and future compatibility with systemd, administrators are encouraged to create own systemd service files or udev rules to run scripts during boot.  This post will briefly outline the use of a udev rule to assign a specific I/O scheduler to a specific HDD.

NOTE To see an example using the phased-out rc-startup scripts, take a look at my previous post.
Create a custom rules-file, e.g. 99-custom.rules, in the /etc/udev/rules.d/ directory with your editor of choice.  The following example will instruct udev to assign the deadline I/O scheduler to the /dev/sdb device:

# vim /etc/udev/rules.d/99-custom.rules ACTION=="add|change", SUBSYSTEM=="block" , KERNEL=="sdb*", RUN+="/bin/sh -c 'echo deadline > /sys/bloc…

Specify I/O Scheduler per Device

Since the advent of systemd the use of /etc/rc* startup-scripts has been discouraged and phased out. However, systemd still supports the use of certain local startup scripts for compatibility purposes. This post will briefly outline the use of the /etc/rc.local file to assign a specific I/O scheduler to a specific HDD.

At the moment, I have an SSD drive (primary /dev/sda) and an "old-fashioned" HDD (/dev/sdb) in use on my RHEL7 system.  I'd like to be able to use the deadline I/O scheduler as the default and assign the cfq scheduler to the HDD device.  In RHEL7 the default I/O scheduler can change based on the selected tuned profile, which adds an additional layer of uncertainty if you're unaware of "tuned".  The default tuned profile is throughput-performance, which enables the deadline scheduler by default among other performance-related system settings.  However, if the default profile is changed to , e.g. virtual-host, the scheduler of choice becomes

SystemD and FIFO Sockets in RHEL7

There's a bug with a relevant discussion on systemd's approach to FIFO socket deletion. As of systemd-214 the issue with "stale" sockets was resolved by supplying the `RemoveOnStop` option to its corresponding `.service`. However, at the moment RHEL7 has systemd-208 as the default version; and I am seeing the following errors in `dmesg` output: systemd[1]: systemd 208 running in system mode. (+PAM +LIBWRAP +AUDIT +SELINUX +IMA +SYSVINIT +LIBCRYPTSETUP +GCRYPT +ACL +XZ) systemd[1]: /usr/lib/systemd/system-generators/anaconda-generator exited with exit status 1. [ +0.056775] systemd[1]: [/usr/lib/systemd/system/lvm2-lvmetad.socket:9] Unknown lvalue 'RemoveOnStop' in section 'Socket' [ +0.000675] systemd[1]: [/usr/lib/systemd/system/dm-event.socket:10] Unknown lvalue 'RemoveOnStop' in section 'Socket' I'm not sure if LVM2 is referencing a feature that's not available in the default systemd version; AFAIK, my configurati…

SystemD Debate and Configuration by Example

Image
A "healthy" debate on the virtues, use-cases, and overall disdain and misconceptions of SystemD. Specific use-cases in "less-common" configurations provide a challenge as well as an opportunity to learn about the underlying mechanism of certain aspects of the Operating System. A substantial number of years working in/with Linux doesn't preclude a person from learning how to use new tools and mechanisms even if, or perhaps especially when, the mechanism is low-level software.

Users and Administrators must not get so entrenched in their ways of comfort that they reject advancements in the underlying pieces of the OS on the basis of overestimating the merit of their own existing knowledge, while downplaying the benefits of the advancements and exaggerating their problems an perceived shortcomings on surface value. Nevertheless, even though change is inevitable in our universe, it breeds chaos in the early stages of wide-spread adaptation. It's in this st…

Interview with SystemD developers

(translated from Golem.de)
Systemd's goal is to standardize Linux, +Lennart Poettering told us three years ago. For quite some time, already, the project that was started by Poettering and +Kay Sievers has become more than simply an init-service; with the upcoming switch to +systemd  by the Debian and Ubuntu Linux distributions, it seems that systemd's goal to standardize Linux has been achieved, according to the developers in the interview with Golem.de.

In the meantime, the only missing OS from among the biggest and most well-known Linux distributions is +Slackware all others have either implemented systemd already, offer a solution for its use, or have begun the changeover to systemd.  That's something that the developers could never have imagined, though did hope; ultimately, they remain convinced of the advantages that their software provides.

In the last four years they were both heavily criticized, though, as Sievers put it, that criticism comes and goe…