On 8/17/2021 01:02, Mario De Chenno wrote:
Hi all. I've been relying on xymon for over ten years, and I would be happy to start it over now that sysadmin is not my core job anymore. Hi,
The project still exists, but the 4.3 branch is definitely in a maintenance/stable mode at this point, pending primarily TLS 1.3 issues that have been reported.
The primary stumbling block to 4.4 (or beyond) really was taking too much to chew in all at once, with IPv6 (which is still rarely necessary at this level) and SSL message validation being the key stumbling blocks (as opposed to SSL-message tunneling, which is the current workaround).
Xymon's function design also naturally tends towards extension at the end of data pipelines rather than the core. This is great for customization as part of a larger work pipeline (with work customizing for local needs), but tends to retard large jumps in the core process. A *lot* of QOL functionality in other monitoring frameworks depends on having a central scheduler and configuration outside of flat files, but the simplicity of explicit design (except for current state with ghosts, which is snapshoted) is also a strength, leaving that part of a bug v feature discussion.
Anyway, while user community is alive and kicking, seems that developing work on the 4.3 and 4.4 branches freezed at some point ?in 2019. Could you project leaders share with us some info on how and where the project is going to?
Alas, on a personal note there've been a number of things that've happened since 2019 (other than the obvious global ones) that forced me to take a step back from the pace I was working on things before... Namely, a loss of local computer equipment (and backups), a flood and rebuild of home (with hotel living for months), and an ill-timed case of Covid. I'd absolutely hoped to have been able to start actual releases again before now.
Also the Devmon fellow project (which AFAIK shares the main developer with Xymon) would need some clarifying words, and merging the efforts instead of forking the project would be a better choice for all xymon users.
SNMP has long been a requested service to be integrated more closely with xymon, both for discovery and state. With devmon persisting so long, there's absolutely a case (in my mind) for closer tying into the core. I know this has been a priority (at least v1/2) for a while, especially at installations where agent capability is limited.
I'm open to suggestions from all on both accounts.
Regards,
-jc