[tickets] [opensaf:tickets] #1925 imm: 2PBE preload information is not sent from standby
- **status**: accepted --> review --- ** [tickets:#1925] imm: 2PBE preload information is not sent from standby** **Status:** review **Milestone:** 5.0.1 **Created:** Thu Jul 21, 2016 04:47 AM UTC by Neelakanta Reddy **Last Updated:** Thu Jul 21, 2016 04:47 AM UTC **Owner:** Neelakanta Reddy In a 2PBE enabled cluster, if both the controllers are started simultaneously, then the IMMNDs in both the controllers will send pre-load information to active IMMD. Based on pr-load information the active IMMD descides form which IMMND the loading should start. with #79 the assignment of standby role is delayed until amfd initialization. The standby IMMND is not considered as a controller IMMND and the preload information is not sent to standby IMMND. solution: Send the preload information to all IMMNDs. If the IMMND is on controller then send the pre-load information. --- Sent from sourceforge.net because opensaf-tickets@lists.sourceforge.net is subscribed to https://sourceforge.net/p/opensaf/tickets/ To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/opensaf/admin/tickets/options. Or, if this is a mailing list, you can unsubscribe from the mailing list.-- What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports.http://sdm.link/zohodev2dev___ Opensaf-tickets mailing list Opensaf-tickets@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-tickets
[tickets] [opensaf:tickets] #1704 imm: Use nanosecond resolution for time measurements
- **status**: unassigned --> accepted - **assigned_to**: Hung Nguyen - **Part**: - --> nd - **Milestone**: future --> 5.1.FC --- ** [tickets:#1704] imm: Use nanosecond resolution for time measurements** **Status:** accepted **Milestone:** 5.1.FC **Created:** Fri Mar 18, 2016 10:08 AM UTC by Anders Widell **Last Updated:** Fri Mar 18, 2016 10:08 AM UTC **Owner:** Hung Nguyen There are currently time measurements for timeout handling in **ImmModel**, and possibly other places in the IMM service, that use low resolution time measurements. This is caused by using the **time_t** type for time stamps, which only has a resolution of seconds. A consequence of this is that the actual timeout can vary up to one second, depending on when (how long after the last clock "tick") the timer is started. Instead, nanosecond resolution should be used. The **osaf_clock_gettime()** utility function declared in *osaftime.h* can be used to read the system clock with nanosecond resolution. Another option in C++ is to use **std::chrono::steady_clock**. This ticket can be seen as a continuation of ticket [#1617], where we started to use **osaf_clock_gettime()** but still didn't change the type of the variables containing time-stamps from **time_t** to **struct timespec**. --- Sent from sourceforge.net because opensaf-tickets@lists.sourceforge.net is subscribed to https://sourceforge.net/p/opensaf/tickets/ To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/opensaf/admin/tickets/options. Or, if this is a mailing list, you can unsubscribe from the mailing list.-- What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports.http://sdm.link/zohodev2dev___ Opensaf-tickets mailing list Opensaf-tickets@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-tickets
[tickets] [opensaf:tickets] #1856 Add IMMSv Readme for Headless feature
- **status**: accepted --> review --- ** [tickets:#1856] Add IMMSv Readme for Headless feature** **Status:** review **Milestone:** 5.0.1 **Created:** Tue May 31, 2016 09:21 AM UTC by Chani Srivastava **Last Updated:** Thu Jul 14, 2016 08:04 AM UTC **Owner:** Hung Nguyen Like other services(amf, log), there is no IMMSv Readme for Headless feature. Currently, readme only describes the SC_ABSENCE feature, but how the immsv service is expected to behave in the absence of controllers is not documented. --- Sent from sourceforge.net because opensaf-tickets@lists.sourceforge.net is subscribed to https://sourceforge.net/p/opensaf/tickets/ To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/opensaf/admin/tickets/options. Or, if this is a mailing list, you can unsubscribe from the mailing list.-- What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports.http://sdm.link/zohodev2dev___ Opensaf-tickets mailing list Opensaf-tickets@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-tickets
[tickets] [opensaf:tickets] #1553 AMF: Send CsiSetCallback immediately when CSI attributes are modified (loadbalancing, VM migration usecases)
Changes at AMFD (CSI Attribute Change Callback): 1)Support for new attribute osafAmfCSICommunicateCsiAttributeChange in class SaAmfCSI. Dynamic modification of this attribute is supported. Default value will be 0. 2)New message AVSV_D2N_COMPCSI_ASSIGN_MSG from AMFD to AMFND for compcsi related information. As of now only CSI attribute list will be sent to AMFND. In future it can be used in the context of tickets like \#538 (TERM_FAILED state) and \#83. 3)Upon modification of CSI attribute value for a object of class "SaAmfCSIAttribute", AMFD will send a message to AMFND with new list. For a NON PROXIED NPI component, message will not be sent if osafAmfCSICommunicateCsiAttributeChange is false. For such a component, change will come into effect only when its SI is locked and unlocked or other such ways. 4)AMFD now also maintains MDS install version of all AMFNDs in std::mapnds_mds_ver_db. It will be updated whenever AMFD gets MDS_UP and MDS_DOWN for AMFND. Using this AMFD can decide whether message is meant for particular AMFND much before encode callback given by MDS. Changes at AMFND (CSI Attribute Change Callback): 1)New internal AMF callback from AMFND to AMF agent corresponding to OsafCsiAttributeChangeCallbackT for giving modified CSI Attribute Name value pairs to component (AVSV_AMF_CSI_ATTR_CHANGE_PARAM). 2)AMFND processes AVSV_D2N_COMPCSI_ASSIGN_MSG from AMFD. Upong receving CSI attribute list, it updates its databae with modified information. If compoent has registered with new callback OsafCsiAttributeChangeCallbackT, then it will send this modification through internal callback AVSV_AMF_CSI_ATTR_CHANGE_PARAM to AMF agent. 3)AMFND now maintains MDS install version of AMF AGENTs in std::map agent_mds_ver_db. It will be updated whenever AMFND gets MDS_UP and MDS_DOWN for AMF Agent. Using this AMFND can decide whether message is meant for particular Agent much before encode callback given by MDS. 4)AMFND maintains now SAF version for each registered component. AMFND will get it from Agent through existing component registeration message. Changes at AMF Agent (CSI Attribute Change Callback): 1)Because of different AMF callback struture correponding to each of B.01.01 B.04.01, introduced an internal callback structure OsafAmfCallbacksT. It consists of callback from each AMF version. Since callback for each intialization needs to be maintained internally in AVA_HDL_REC. With this internal structure, same handle can used internally for any initialization of AMF service. Here it is: typedef struct osaf_amf_cbk { SaAmfHealthcheckCallbackT saAmfHealthcheckCallback; SaAmfComponentTerminateCallbackT saAmfComponentTerminateCallback; SaAmfCSISetCallbackTsaAmfCSISetCallback; SaAmfCSIRemoveCallbackT saAmfCSIRemoveCallback; SaAmfProtectionGroupTrackCallbackT saAmfProtectionGroupTrackCallback; SaAmfProtectionGroupTrackCallbackT_4 saAmfProtectionGroupTrackCallback_4; //Only in B.04.01 SaAmfProxiedComponentInstantiateCallbackT saAmfProxiedComponentInstantiateCallback; SaAmfProxiedComponentCleanupCallbackT saAmfProxiedComponentCleanupCallback; SaAmfContainedComponentInstantiateCallbackT saAmfContainedComponentInstantiateCallback; SaAmfContainedComponentCleanupCallbackT saAmfContainedComponentCleanupCallback; OsafCsiAttributeChangeCallbackT osafCsiAttributeChangeCallback; //OpenSAF implementation. } OsafAmfCallbacksT; Also added utility function to copy callbacks from AMF callbacks structure to this internal one. AMF agent will use OsafAmfCallbacksT internally instead of SaAmfCallbacksT_<#x>. 2)Implementation of saAmfInitialize_5() in ava_api.cc. 3)Now saAmfRegister() also sends SAF version to AMFND. Will update more on AMFD and AMFND MDS versioning before publishing the patches. --- ** [tickets:#1553] AMF: Send CsiSetCallback immediately when CSI attributes are modified (loadbalancing, VM migration usecases)** **Status:** accepted **Milestone:** 5.1.FC **Labels:** VM migration loadbalancer-synchronization **Created:** Wed Oct 21, 2015 10:03 AM UTC by Mathi Naickan **Last Updated:** Mon Jul 25, 2016 06:09 AM UTC **Owner:** Praveen AMF is the 'workload' manager in OpenSAF. There are also loadbalancers external to OpenSAF that can choose to modify some attributes of an application component's workload which is currently assigned ACTIVE (without loss of service). In such scenarios it is possible (and necessary) for both AMF and the external load balancers to throttle each other. The reverse throttling(from AMF to the external load balancer) is
[tickets] [opensaf:tickets] #1454 amf: Support N+M model for shutdown operation on node group
High Level Changes at AMFD: \-Currently nodegroup.cc contains all checks related to validityof admin operation on NG hosting SU of a given red model. These checks are applicable to sg of N+M model also. There is almost no change in this file except the removal of check on red model. \-After checking validity of lock and shutdown operation, AMFD calls SG FSM function for assignment handling for each SU hosted on a given node from NG's node list. Thus a new function ng_admin() is added in sg_npm_fsm.cc. This new function works at su level. Depending upon the HA state of SU, assignment message is sent to the Node director that hosts this SU. If SU has active state then quiescing or quiesced HA state is sent respectively for shutdown and lock operation. If SU has standby HA state then removal is sent for both shutdown and lock operations. SG is moved to REALIGN state. Upon receiving successful response for quiescing or quiesced assignements, AMFD tries to switchover this SU by making its peer SU active if SU is not assigned on nodes of this nodegroup. If switchover of whole SU is not possbile then AMFD tries to make as many SUSI active as possible and removing other SUSI from the peer SU. In case of successful removal o assignments, SUSIs are deleted at AMFD. In case of shutdown admin op, initial admin state of NG and Node will be SHUTTING_DOWN. When all the SUs of a node moves to quiesced state then Node transitions to LOCKED state. A NG transitions to LOCKED state when all its nodes transitions to LOCKED state. Admin operation finishes when all assignments gets removed in each SU hosted on nodes of NG. Since SG is moved to REALIGN state, any faults occured during assignment is handled by re-using red-model specific fault related code functions. \-For lock-in and unlock-in admin operations, there are no new changes execpt the removal of check on red models. In case of unlock-in operation, SUs are instantiated honoring PrefInserviceSU criteria. \-For unlock operations, SUs are assigned based on their ranks in SG on the nodes of NG. This is currenlty done by calling su_insvc() function from nodegroup.cc. There are also no new changes in this area execpt removal of red model check. Note: There are no changes at AMFND and AMFA. --- ** [tickets:#1454] amf: Support N+M model for shutdown operation on node group** **Status:** review **Milestone:** 5.1.FC **Created:** Fri Aug 14, 2015 02:07 PM UTC by Anders Widell **Last Updated:** Fri Jul 22, 2016 05:24 AM UTC **Owner:** Praveen This ticket aims to support N+M model for: 1)shutdown, 2) lock, 3) lock-in and 4) unlock-in operations on nodegroup. Note: SI deps will not be honored, within SU, while giving quiesced/quiescing assigments. --- Sent from sourceforge.net because opensaf-tickets@lists.sourceforge.net is subscribed to https://sourceforge.net/p/opensaf/tickets/ To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/opensaf/admin/tickets/options. Or, if this is a mailing list, you can unsubscribe from the mailing list.-- What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports.http://sdm.link/zohodev2dev___ Opensaf-tickets mailing list Opensaf-tickets@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-tickets
[tickets] [opensaf:tickets] #1608 amf: Support N_WAY model for admin operations on node group.
High Level Changes at AMFD: \-Currently nodegroup.cc contains all checks related to validityof admin operation on NG hosting SU of a given red model. These checks are applicable to sg of N-Way model also. There is almost no change in this file except the removal of check on red model. \-After checking validity of lock and shutdown operation, AMFD calls SG FSM function for assignment handling for each SU hosted on a given node from NG's node list. Thus a new function ng_admin() is added in sg_nway_fsm.cc. This new function works at su level. Depending upon the HA state of SU, assignment message is sent to the Node director that hosts this SU. If SU has active state then quiescing or quiesced HA state is sent respectively for shutdown and lock operation. If SU has standby HA state then removal is sent for both shutdown and lock operations. SG is moved to REALIGN state. Upon receiving successful response for quiescing or quiesced assignements, AMFD tries to switchover each SUSI of this SU by making its peer SUSI active if they are not assigned on nodes of this nodegroup. If switchover is not possbile then AMFD sends removal of assignment to Node director hosting this quiescing/quiesced SU. After successful removal, SUSI are deleted at AMFD. In case of shutdown adm in op, initial admin state of NG and Node will be SHUTTING_DOWN. When all the SUs of a node moves to quiesced state then Node transitions to LOCKED state. A NG transitions to LOCKED state when all its nodes transitions to LOCKED state. Admin operation finishes when all assignments gets removed in each SU hosted on nodes of NG. Since SG is moved to REALIGN state, any faults occured during assignment is handled by re-using red-model specific fault related code functions. \-For lock-in and unlock-in admin operations, there are no new changes execpt the removal of check on red models. In case of unlock-in operation, SUs are instantiated honoring PrefInserviceSU criteria. \-For unlock operations, SUs are assigned based on their ranks in SG on the nodes of NG. This is currenlty done by calling su_insvc() function from nodegroup.cc. There are also no changes in this area execpt removal of red model check. Note: There are no changes at AMFND and AMFA. --- ** [tickets:#1608] amf: Support N_WAY model for admin operations on node group.** **Status:** review **Milestone:** 5.1.FC **Created:** Tue Nov 24, 2015 06:55 AM UTC by Praveen **Last Updated:** Fri Jun 03, 2016 10:19 AM UTC **Owner:** Praveen Currently AMF supports 2N, NoRed and Nway_active models for 1)shutdown, 2) lock, 3) lock-in and 4) unlock-in operations on nodegroup. This ticket is for supporting N_WAY model for above mentioned operations on Node Group. Note: SI deps will not be honored, within SU, while giving quiesced/quiescing assigments. --- Sent from sourceforge.net because opensaf-tickets@lists.sourceforge.net is subscribed to https://sourceforge.net/p/opensaf/tickets/ To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/opensaf/admin/tickets/options. Or, if this is a mailing list, you can unsubscribe from the mailing list.-- What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic patterns at an interface-level. Reveals which users, apps, and protocols are consuming the most bandwidth. Provides multi-vendor support for NetFlow, J-Flow, sFlow and other flows. Make informed decisions using capacity planning reports.http://sdm.link/zohodev2dev___ Opensaf-tickets mailing list Opensaf-tickets@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-tickets