In a perfect world I would prefer for a downed slave to not block many actions in Opsview. I would think the Opsview master server would have a warning status on the load configuration screen to let me know a config update or slave upgrade did not complete. There would be a pending actions queue that could be re-run to clear the warning status.
In our case we have a lot of slave servers at remote locations so the blocking behavior of applying a config is the archilles heel so to say, at least from the service provider perspective. I understand there are many valid reasons to keep everything in synch but I would think the architecture could be tweaked to be more accepting of downed monitoring nodes. James Whittington VC3, Inc. From: [email protected] [mailto:[email protected]] On Behalf Of Ton Voon Sent: Thursday, December 09, 2010 9:10 AM To: Opsview Users Subject: Re: [opsview-users] FW: How do we push new versions to the slave servers after the master is upgraded? Thanks for all the feedback. Emilio, that's a good idea re: parameter to fail the upgrade or not, so I've updated the description at https://secure.opsera.com/jira/browse/OPS-896 Paul, I've noted the idea re: different ssh ports in https://secure.opsera.com/jira/browse/OPS-1455 Ton On 9 Dec 2010, at 12:33, Emilio Scalise wrote: When I upgrade opsview I often need to disable my slave machine and then re-enable later it. It would be nice if there is an option to disable the fail if there is an error upgrading slaves behaviour (naturally the default setting should be the current behaviour). Regards, Emilio Il 09/12/2010 12:49, Ton Voon ha scritto: On 8 Dec 2010, at 17:53, Paul wrote: Hi James, ASAIK the suggested way you describe below should work OK. I'm running slaves on a different port than 22, so upgrade always yell slaves can;t be updated. Then I modify two files, indicating the correct port and re-run the update process - which has worked for me for the past 5 times :) Good luck Paul Hi Paul, One thing we were thinking was to force an upgrade to fail if the slaves failed to update. This is the most common case and necessary if, for instance, we make a major update to nagios which will generate a new style configuration which an older nagios would die trying to run (thus slaves just appear to have stopped working) However, this blocks your process. Suggestions? Ton _______________________________________________ Opsview-users mailing list [email protected]<mailto:[email protected]> http://lists.opsview.org/lists/listinfo/opsview-users <ATT00001..c> Ton Voon Product Architect [cid:[email protected]] Global Headquarters: Unit 69 Suttons Business Park | Reading | Berkshire | RG6 1AZ | UK UK: +44 (0) 845 057 7887 USA: +1 866 879 9184 Email: [email protected]<mailto:[email protected]> Skype: tonvoon Web: www.opsview.com<http://www.opsview.com/> This e-mail is confidential, intended only for the named recipient(s) above and may contain information that is privileged and confidential. If you receive this message in error, or are not the named recipient(s), please notify the sender at the phone number above, do not copy this message, do not disclose its contents to anyone, and delete this e-mail message from your computer. Although we routinely screens for viruses, addressees should scan this e-mail and any attachments for viruses. Opsera makes no representation or warranty as to the absence of viruses in this e-mail or any attachments.
<<inline: image001.png>>
_______________________________________________ Opsview-users mailing list [email protected] http://lists.opsview.org/lists/listinfo/opsview-users
