PROTECTION OF I/O PATHS AGAINST NETWORK PARTITIONING AND COMPONENT FAILURES IN NVME-OF ENVIRONMENTS
A Centralized Discovery Controller (CDC) uses built-in intelligence to determine or estimate whether a connection with a non-volatile memory express (NVMe) entity has been discontinued intentionally or whether a connection loss is rather transient, e.g., due to a temporary network issue. In the former case, the CDC sends out asynchronous event notifications (AENs) and communicates the absence of the NVMe entity in a get log page to indicate an administrative access control action, e.g., a user intervention due to a zoning change or the removal of an entity due to a hardware failure. In the latter case, the CDC creates an “unreachable” entry in the name server database that indicates a CDC connectivity failure but maintains the entity in the name server database despite the connection loss, refraining from sending out notifications to relevant (or impacted) entities to increase bandwidth, traffic stability, and network availability.
1 . A computer-implemented method for increasing bandwidth and network availability in a storage area network (SAN), the computer-implemented method comprising:
in response to a Centralized Discovery Controller (CDC) in a SAN determining a connection loss between the CDC and a first non-volatile memory express (NVMe) entity, generating a notification that indicates that the CDC will not remove the first NVMe entity from a database despite the connection loss;
communicating the notification to a second NVMe entity causing the second NVMe entity to not disconnect from the first NVMe entity, thereby, improving traffic stability; and
in response to the CDC determining a purging condition for the first NVMe entity, removing the first NVMe entity from the database, such that a query response, made by the CDC in response to a query by the second NVMe entity, does not contain the first NVMe entity.
2 . The computer-implemented method of claim 1 , wherein the first NVMe entity is a host, the second NVMe entity is a subsystem, the database is a name server database maintained in the CDC, and the notification is at least one of an entry or a flag in the database.
3 . The computer-implemented method of claim 1 , wherein the notification further indicates that the first NVMe entity is temporarily unreachable due to the connection loss.
4 . The computer-implemented method of claim 1 , further comprising in response to determining the connection loss, not communicating an asynchronous event notification (AEN) to the second NVMe entity to prevent soliciting the query.
5 . The computer-implemented method of claim 1 , wherein removing the first NVMe entity from the database is performed by a networking component different from the CDC.
6 . The computer-implemented method of claim 5 , wherein removing is performed according at least one of a policy, a configuration setting, or a maintenance procedure.
7 . The computer-implemented method of claim 1 , further comprising in response to determining the purging condition, overriding the notification by sending out an asynchronous event notification (AEN) to the second NVMe entity.
8 . The computer-implemented method of claim 1 , wherein the purging condition comprises at least one of:
a termination request by the first NVMe entity,
a lack of communication by the first NVMe entity for a period of time, or
detecting by the CDC at least one of:
a name server move,
a replacement of the first NVMe entity on a physical switch port,
a forced removal of the first NVMe entity from a name server database,
a deletion of an NVMe entity reference from a zone database, or
a deletion of the first NVMe entity based on a time out condition.
9 . A non-transitory computer-readable medium or media comprising one or more sequences of instructions which, when executed by at least one processor, causes steps to be performed comprising:
in response to receiving from a Centralized Discovery Controller (CDC) a notification that indicates a connection loss between the CDC and a non-volatile memory express (NVMe) entity but that the CDC will not remove the NVMe entity from its database, not terminating a connection with the NVMe entity to improve traffic stability;
in response to the CDC determining a purging condition for the NVMe entity and the NVMe entity being removed from the database, receiving from the CDC an asynchronous event notification (AEN);
sending a query to the CDC;
receiving a from the CDC a query response that does not contain the NVMe entity; and
terminating the connection with the NVMe entity.
10 . The non-transitory computer-readable medium or media of claim 9 , wherein the CDC removes the NVMe entity from database by performing at least one of unflagging or changing a bit in a name server database.
11 . The non-transitory computer-readable medium or media of claim 9 , wherein the NVMe entity is removed from the database, according at least one of a policy, a configuration setting, or a maintenance procedure, by a networking component different from the CDC.
12 . The non-transitory computer-readable medium or media of claim 9 , wherein the notification further indicates that the NVMe entity is temporarily unreachable due to a connectivity failure.
13 . The non-transitory computer-readable medium or media of claim 9 , further comprising not receiving an AEN in response to determining the connection loss.
14 . The non-transitory computer-readable medium or media of claim 9 , further comprising receiving an AEN in response to the CDC determining the purging condition.
15 . The non-transitory computer-readable medium or media of claim 9 , wherein the purging condition comprises at least one of:
a termination request by the first NVMe entity,
a lack of communication by the first NVMe entity for a period of time, or
detecting by the CDC at least one of:
a name server move,
a replacement of the first NVMe entity on a physical switch port,
a forced removal of the first NVMe entity from a name server database,
a deletion of an NVMe entity reference from a zone database, or
a deletion of the first NVMe entity based on a time out condition.
16 . A system for increasing bandwidth and network availability in a storage area network, the system comprising:
one or more processors; and
a non-transitory computer-readable medium or media comprising one or more sets of instructions which, when executed by at least one of the one or more processors, causes steps to be performed comprising:
in response to a Centralized Discovery Controller (CDC) in a SAN determining a connection loss between the CDC and a first non-volatile memory express (NVMe) entity, generating a notification that indicates that the CDC will not remove the first NVMe entity from a database despite the connection loss;
communicating the notification to a second NVMe entity causing the second NVMe entity to not disconnect from the first NVMe entity, thereby, improving traffic stability; and
in response to the CDC determining a purging condition for the first NVMe entity, removing the first NVMe entity from the database, such that a query response, made by the CDC in response to a query by the second NVMe entity, does not contain the first NVMe entity.
17 . The system of claim 16 , wherein the notification further indicates that the first NVMe entity is temporarily unreachable due to a connectivity failure.
18 . The system of claim 16 , wherein a networking component different from the CDC performs removing according at least one of a policy, a configuration setting, or a maintenance procedure.
19 . The system of claim 16 , further comprising in response to determining the connection loss not communicating an asynchronous event notification (AEN) to the second NVMe entity to prevent soliciting a query from the second NVMe entity.
20 . The system of claim 16 , further comprising in response to determining the purging condition, overriding the notification by sending out an asynchronous event notification (AEN) to the second NVMe entity.