Linux Collectors: Expanded OS Support
We recently expanded our list of supported Linux distributions by adding Oracle, Chainguard, and Rocky Linux. Also, at the end of June we'll no longer support CentOS, as this distribution is end of life and no longer maintained. https://www.logicmonitor.com/support/adding-collector#h-collector-server-requirements Excerpt support list for Collector OS: Amazon Linux CentOS Linux (LogicMonitor End of Support: July 1, 2026) Chainguard Debian GNU/Linux Oracle Linux Server Red Hat Enterprise Linux (RHEL) Rocky Linux Ubuntu With this, over 93% of all Linux collectors used by our customers are using a LogicMonitor-supported Linux distribution. However, a little south of 5% of Linux collectors are still running CentOS that reached end-of-life four years ago, so if this applies to please make plans to migrate to a supported Linux distro to mitigate risk of monitoring disruption and to maintain a strong security posture.
CentOS IP Tables?
We don't do much monitoring of servers at the moment, but I received an inquiry about the possibility of monitoring iptables data for a CentOS server. I've looked through modules in the exchange and I'm not seeing much that seems to do this but I was curious if anyone else has ever looked into this. I'm not a server person so I am also not as knowledgeable about how this works as I'd like to be Thank youLinux details for EC2 hosts
Linux details for EC2 hosts such as Kernel version, etc are not displayed, as the Linux_SSH datasource specifically excludes these. I can understand that for some hardware specific properties/info, but it would be useful to have the patch release, etc. of the kernel regardless of being a VM/EC2 in AWS. Is there another datasource or property source I should be using for this info?Migrate your Linux collectors to non-root by Sept 30!
Hello All, Thank you for supporting LogicMonitor's efforts to ensure Collector Security. With your help, we have been able to transition ~7,000 collectors to non-privileged users out of the 10,000 linux collectors currently live in customer environments. Per our last email on this topic, we had shared a deadline of June 30, 2023 for customers to migrate their collectors from root to non-root users. Due to customer requests needing more time, we have now extended the deadline to September 30, 2023, allowing for more time to test the non-root migration scripts and migrate linux collectors. We appreciate your support in helping us achieve our goal of running all collectors using non-privileged credentials. ACTION REQUIRED: Migrate any collectors which are running under root users to non-privileged users For more details, please refer to: https://www.logicmonitor.com/support/migrating-collector-from-root-to-non-root-user If your current collector installation process uses the root user to install linux collectors, please start using non-privileged user For more details, please refer to: https://www.logicmonitor.com/support/collectors/collector-installation/installing-collectors#Linux-collector. TIMELINE: Migrate your current linux collector install base to non-privileged users as soon as possible, however no later than September 30, 2023. Your current collectors will not be affected by this change, only new installs will not be installed as root. Thank you for your prompt attention to this matter. If you have any questions, please contact Logicmonitor Support or contact your Customer Success Manager (CSM) Is there a reason why this will not work in your environment? Would you still like to run a linux collector as root? Let us know in the comments Thank you!
eBPF monitoring support for Linux hosts
I have seen this with some other enterprise solutions and it’s a popular choice when we don’t want ‘APM’ level metrics but we are interested in the HTTP metrics such as requests/sec, error/sec, etc. Redhat has a decent article on this: Monitoring eBPF-based metrics (redhat.com) Some blogs focussed on HTTP metrics: eBPF enhanced HTTP observability — L7 metrics and tracing | by Apache SkyWalking | Jan, 2023 | Medium I am starting to see DevOps requests this sort of monitoring especially on third party software where instrumentation for APM is not possible or preferred by vendor. We’d recently had a requirement similar to this for an enterprise product and LM can only poll HTTP and record results and not be able to track incoming requests to the server. Quite like to see this being supported - DataDog already supports this and I can see this being quite important going forward.