Hi José Tomás Tocino García, System controllers have cluster wide information and they are modeled in 2N redundancy(Active/Standby) to protect the single point of failure. In your case, you are stopping both the controllers. In such cases, to avoid the data inconsistencies, payloads have been designed to reboot.
But, there has been an amazing feature added later on, called headless feature or 'absence of system controllers', which can avoid payloads from rebooting. In case, both the controllers go down, the payloads stands still. When controllers comes back, it takes necessary information from payloads and starts functioning again. Till controllers are down, payloads may not be able to take cluster wide decisions and hence they defer their decisions till at least one controller comes back. You can read "6 Absence of System Controllers" of OpenSAF_Extensions_PR.odt available at https://sourceforge.net/p/opensaf/documentation/ci/default/tree/ Hope it helps! Have a great day ahead! Thanks & Best Regards -Nagendra, +91-9866424860 www.GetHighAvailability.com Get High Availability Today! NJ, USA: +1 508-422-7725 | Hyderabad, India: +91 798-992-5293 -----Original Message----- From: Tocino García, José Tomás [ELIMCO] [mailto:elimco.jttoci...@navantia.es] Sent: 10 February 2020 18:56 To: opensaf-users@lists.sourceforge.net Subject: [users] Avoid payload node automatic reboot Hello. I'm slowly wrapping my head around OpenSAF. I'm currently running a 4-node (2 SC, 2 PL), 2N cluster with the basic AmfDemo application, and an external development workstation from which I make the changes to the config files that are later scp'd into the cluster. If I restart OpenSaf on the two System Controllers, suddenly (and understandably) the Payload nodes reboot automatically, launching this log: Feb 10 13:01:01 proc0105 osafamfnd[2908]: WA AMF director unexpectedly crashed Feb 10 13:01:01 proc0105 osafamfnd[2908]: Rebooting OpenSAF NodeId = 140047 EE Name = , Reason: local AVD down(Adest) or both AVD down(Vdest) received, OwnNodeId = 140047, SupervisionTime = 60 I have two questions. First of all, where can I read more about this behavior? As far as I've read, it looks like it's performing an SA_AMF_NODE_FAILFAST, but there's nothing related to it in the logs. The second question is, how can I prevent this behavior? I'm guessing the payload nodes cannot stay active without any SystemController, but it would be great if they didn't reboot the hardware. The XML are barely modified, I generated the base xml files from the nodes.cfg and then merged the AppConfig-2N.xml from the AmfDemo file. Thanks. -- José Tomás Tocino García Ingeniero Informático - System Infrastructure Team Ubicación: Edif. Integración LBTS F110 / F105 / SCOMBA, Navantia Sistemas, SF Email: elimco.jttoci...@navantia.es<mailto:elimco.jttoci...@navantia.es> Tfno: 856 30 9163 [logoSoologicSmall] [Navantia] ________________________________ NAVANTIA S.A. S.M.E. Este mensaje y cualquier fichero anexo al mismo contiene información de carácter confidencial dirigida exclusivamente a su(s) destinatario(s) y, en su caso, sometida a secreto profesional. Queda prohibida su difusión, copia o distribución a terceros sin la previa autorización escrita. Si Vd. ha recibido este mensaje por error, se ruega lo comunique inmediatamente por esta misma vía y proceda a su completa eliminación. Puede revisar nuestra política de privacidad en http://www.navantia.es/es/legal/. The information in this e-mail and in any attachments is confidential and, if any, protected by a professional privilege and intended solely for the attention and use of the named address(es). You are hereby notified that any dissemination, copy or distribution of this information is prohibited without the prior written consent. If you have received this communication in error, please notify the sender by reply e-mail and delete it. You can review our privacy policy at http://www.navantia.es/en/legal/. ________________________________ [Navantia] Piense en el medio ambiente. ¿Necesita realmente imprimir este correo? Please care for the environment. Do you really need to print this e-mail? _______________________________________________ Opensaf-users mailing list Opensaf-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-users _______________________________________________ Opensaf-users mailing list Opensaf-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-users
