Comware

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

I see. Just FYI, Comware 7 version numeration can be tricky - sometimes 2609 for one platform is older and has less features than 2432 for another platform. Or 2609 for one model is not the same as 2609 for another model. In other words in Comware world there is no 'single image' paradigm many other vendors brag about :) 

"Line protocol state: Down(LAGG)" means that local physical port's properties or flags that it's got from its peer are different from the properties or flags first port have. The issue can be either in LACP protocol or in cables, since you mentioned 'custom-made DACs'. In order to understand what issue has been resolved in 2609 for 5940 concerning aggregated links, could you send me the link to that document?

TBH there is one thing that 'bothers' me in the network topology you are trying to build. You say "I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches." Those 5900AF, are they in IRF?
Another thing - "Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch". Even if 5900AF switches are in IRF (which means they both look like single physical device to neighboring hosts), what about those two Dell switches? Do they form a stack? If they don't, do they have any kind of MLAG to present themselves to the 5900AF IRF as one single LACP peer? For now it looks like two Dell switches work independently and present themselves with different SystemIDs to 5900 stack, therefore LACP process thinks it's a wiring error and doesn't allow second port to join the BAGG. In order this setup to work you need to make both Dell switches to present themselves with one common LACP SystemID to our stack, Dell calls this technology 'MLAG'.

It will help if you post an output of 'display link-aggregation verbose' in particular its section concerning that non-working BAGG. Then I will tell you for sure if my assumption about lack of MLAG on Dell's side is correct.

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

I see. Just FYI, Comware 7 version numeration can be tricky - sometimes 2609 for one platform is older and has less features than 2432 for another platform. Or 2609 for one model is not the same as 2609 for another model. In other words in Comware world there is no 'single image' paradigm many other vendors brag about :) 

"Line protocol state: Down(LAGG)" means that local physical port's properties or flags that it's got from its peer are different from the properties or flags first port have. The issue can be either in LACP protocol or in cables, since you mentioned 'custom-made DACs'. In order to understand what issue has been resolved in 2609 for 5940 concerning aggregated links, could you send me the link to that document?

TBH there is one thing that 'bothers' me in the network topology you are trying to build. You say "I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches." Those 5900AF, are they in IRF?
Another thing - "Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch". Even if 5900AF switches are in IRF (which means they both look like single physical device to neighboring hosts), what about those two Dell switches? Do they form a stack? If they don't, do they have any kind of MLAG to present themselves to the 5900AF IRF as one single LACP peer? For now it looks like two Dell switches work independently and present themselves with different SystemIDs to 5900 stack, therefore LACP process thinks it's a wiring error and doesn't allow second port to join the BAGG. In order this setup to work you need to make both Dell switches to present themselves with one common LACP SystemID to our stack, Dell calls this technology 'MLAG'.

It will help if you post an output of 'display link-aggregation verbose' in particular its section concerning that non-working BAGG. Then I will tell you for sure if my assumption about lack of MLAG on Dell's side is correct.

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

I see. Just FYI, Comware 7 version numeration can be tricky - sometimes 2609 for one platform is older and has less features than 2432 for another platform. Or 2609 for one model is not the same as 2609 for another model. In other words in Comware world there is no 'single image' paradigm many other vendors brag about :) 

"Line protocol state: Down(LAGG)" means that local physical port's properties or flags that it's got from its peer are different from the properties or flags first port have. The issue can be either in LACP protocol or in cables, since you mentioned 'custom-made DACs'. In order to understand what issue has been resolved in 2609 for 5940 concerning aggregated links, could you send me the link to that document?

TBH there is one thing that 'bothers' me in the network topology you are trying to build. You say "I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches." Those 5900AF, are they in IRF?
Another thing - "Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch". Even if 5900AF switches are in IRF (which means they both look like single physical device to neighboring hosts), what about those two Dell switches? Do they form a stack? If they don't, do they have any kind of MLAG to present themselves to the 5900AF IRF as one single LACP peer? For now it looks like two Dell switches work independently and present themselves with different SystemIDs to 5900 stack, therefore LACP process thinks it's a wiring error and doesn't allow second port to join the BAGG. In order this setup to work you need to make both Dell switches to present themselves with one common LACP SystemID to our stack, Dell calls this technology 'MLAG'.

It will help if you post an output of 'display link-aggregation verbose' in particular its section concerning that non-working BAGG. Then I will tell you for sure if my assumption about lack of MLAG on Dell's side is correct.

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

Ok, now it's clear:

"Aggregate Interface: Bridge-Aggregation30Aggregation Mode: DynamicLoadsharing Type: SharManagement VLAN : NoneSystem ID: 0x8000, 2c23-3a3b-1060Local:Port Status Priority Oper-Key Flag--------------------------------------------------------------------------------XGE1/0/40 S 32768 9 {ACDEFG}XGE1/0/41 U 32768 9 {ACG}Remote:Actor Partner Priority Oper-Key SystemID Flag--------------------------------------------------------------------------------XGE1/0/40 0 32768 0 0x8000, 0000-0000-0000 {DEF}XGE1/0/41 0 32768 0 0x8000, 0000-0000-0000 {DEF}

If you look at the 'SystemID' column of Remote section, you will see all zeroes. Those mean that neither of Dell switches sends LACP packets. Otherwise, even if they don't use MLAG, but use LACP each Dell will present itself with some ID, a number often equals to MAC address of the chassis. In your case LACP is not used at all on Dell's side. The reason one port is UP is simple - Comware keeps BAGG reference port (one of the ports in the BAGG which settings are reference for other ports). But it doesn't mean BAGG is operational over that port, it's not. Ten1/0/40 works more like a standalone port at this moment. 

So, two options:
- If Ten1/0/40 and Ten1/0/41 are connected to one physical Dell switch, then Dell's LAG doesn't use LACP, so you need to enable it there and everything should start working.
- If Ten1/0/40 and Ten1/0/41 are connected to two physical Dell switches, then check this link https://downloads.dell.com/manuals/common/powerconnect_nx00-usingmlag-whtppr_en-us.pdf Your Dells should form "L2 MLAG Domain" and our 5900AF stack will be on the place of 'Partner swtich':

I see. Just FYI, Comware 7 version numeration can be tricky - sometimes 2609 for one platform is older and has less features than 2432 for another platform. Or 2609 for one model is not the same as 2609 for another model. In other words in Comware world there is no 'single image' paradigm many other vendors brag about :) 

"Line protocol state: Down(LAGG)" means that local physical port's properties or flags that it's got from its peer are different from the properties or flags first port have. The issue can be either in LACP protocol or in cables, since you mentioned 'custom-made DACs'. In order to understand what issue has been resolved in 2609 for 5940 concerning aggregated links, could you send me the link to that document?

TBH there is one thing that 'bothers' me in the network topology you are trying to build. You say "I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches." Those 5900AF, are they in IRF?
Another thing - "Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch". Even if 5900AF switches are in IRF (which means they both look like single physical device to neighboring hosts), what about those two Dell switches? Do they form a stack? If they don't, do they have any kind of MLAG to present themselves to the 5900AF IRF as one single LACP peer? For now it looks like two Dell switches work independently and present themselves with different SystemIDs to 5900 stack, therefore LACP process thinks it's a wiring error and doesn't allow second port to join the BAGG. In order this setup to work you need to make both Dell switches to present themselves with one common LACP SystemID to our stack, Dell calls this technology 'MLAG'.

It will help if you post an output of 'display link-aggregation verbose' in particular its section concerning that non-working BAGG. Then I will tell you for sure if my assumption about lack of MLAG on Dell's side is correct.

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

Ok, now it's clear:

"Aggregate Interface: Bridge-Aggregation30Aggregation Mode: DynamicLoadsharing Type: SharManagement VLAN : NoneSystem ID: 0x8000, 2c23-3a3b-1060Local:Port Status Priority Oper-Key Flag--------------------------------------------------------------------------------XGE1/0/40 S 32768 9 {ACDEFG}XGE1/0/41 U 32768 9 {ACG}Remote:Actor Partner Priority Oper-Key SystemID Flag--------------------------------------------------------------------------------XGE1/0/40 0 32768 0 0x8000, 0000-0000-0000 {DEF}XGE1/0/41 0 32768 0 0x8000, 0000-0000-0000 {DEF}

If you look at the 'SystemID' column of Remote section, you will see all zeroes. Those mean that neither of Dell switches sends LACP packets. Otherwise, even if they don't use MLAG, but use LACP each Dell will present itself with some ID, a number often equals to MAC address of the chassis. In your case LACP is not used at all on Dell's side. The reason one port is UP is simple - Comware keeps BAGG reference port (one of the ports in the BAGG which settings are reference for other ports). But it doesn't mean BAGG is operational over that port, it's not. Ten1/0/40 works more like a standalone port at this moment. 

So, two options:
- If Ten1/0/40 and Ten1/0/41 are connected to one physical Dell switch, then Dell's LAG doesn't use LACP, so you need to enable it there and everything should start working.
- If Ten1/0/40 and Ten1/0/41 are connected to two physical Dell switches, then check this link https://downloads.dell.com/manuals/common/powerconnect_nx00-usingmlag-whtppr_en-us.pdf Your Dells should form "L2 MLAG Domain" and our 5900AF stack will be on the place of 'Partner swtich':

I see. Just FYI, Comware 7 version numeration can be tricky - sometimes 2609 for one platform is older and has less features than 2432 for another platform. Or 2609 for one model is not the same as 2609 for another model. In other words in Comware world there is no 'single image' paradigm many other vendors brag about :) 

"Line protocol state: Down(LAGG)" means that local physical port's properties or flags that it's got from its peer are different from the properties or flags first port have. The issue can be either in LACP protocol or in cables, since you mentioned 'custom-made DACs'. In order to understand what issue has been resolved in 2609 for 5940 concerning aggregated links, could you send me the link to that document?

TBH there is one thing that 'bothers' me in the network topology you are trying to build. You say "I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches." Those 5900AF, are they in IRF?
Another thing - "Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch". Even if 5900AF switches are in IRF (which means they both look like single physical device to neighboring hosts), what about those two Dell switches? Do they form a stack? If they don't, do they have any kind of MLAG to present themselves to the 5900AF IRF as one single LACP peer? For now it looks like two Dell switches work independently and present themselves with different SystemIDs to 5900 stack, therefore LACP process thinks it's a wiring error and doesn't allow second port to join the BAGG. In order this setup to work you need to make both Dell switches to present themselves with one common LACP SystemID to our stack, Dell calls this technology 'MLAG'.

It will help if you post an output of 'display link-aggregation verbose' in particular its section concerning that non-working BAGG. Then I will tell you for sure if my assumption about lack of MLAG on Dell's side is correct.

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks

Ok, now it's clear:

"Aggregate Interface: Bridge-Aggregation30Aggregation Mode: DynamicLoadsharing Type: SharManagement VLAN : NoneSystem ID: 0x8000, 2c23-3a3b-1060Local:Port Status Priority Oper-Key Flag--------------------------------------------------------------------------------XGE1/0/40 S 32768 9 {ACDEFG}XGE1/0/41 U 32768 9 {ACG}Remote:Actor Partner Priority Oper-Key SystemID Flag--------------------------------------------------------------------------------XGE1/0/40 0 32768 0 0x8000, 0000-0000-0000 {DEF}XGE1/0/41 0 32768 0 0x8000, 0000-0000-0000 {DEF}

If you look at the 'SystemID' column of Remote section, you will see all zeroes. Those mean that neither of Dell switches sends LACP packets. Otherwise, even if they don't use MLAG, but use LACP each Dell will present itself with some ID, a number often equals to MAC address of the chassis. In your case LACP is not used at all on Dell's side. The reason one port is UP is simple - Comware keeps BAGG reference port (one of the ports in the BAGG which settings are reference for other ports). But it doesn't mean BAGG is operational over that port, it's not. Ten1/0/40 works more like a standalone port at this moment. 

So, two options:
- If Ten1/0/40 and Ten1/0/41 are connected to one physical Dell switch, then Dell's LAG doesn't use LACP, so you need to enable it there and everything should start working.
- If Ten1/0/40 and Ten1/0/41 are connected to two physical Dell switches, then check this link https://downloads.dell.com/manuals/common/powerconnect_nx00-usingmlag-whtppr_en-us.pdf Your Dells should form "L2 MLAG Domain" and our 5900AF stack will be on the place of 'Partner swtich':

I see. Just FYI, Comware 7 version numeration can be tricky - sometimes 2609 for one platform is older and has less features than 2432 for another platform. Or 2609 for one model is not the same as 2609 for another model. In other words in Comware world there is no 'single image' paradigm many other vendors brag about :) 

"Line protocol state: Down(LAGG)" means that local physical port's properties or flags that it's got from its peer are different from the properties or flags first port have. The issue can be either in LACP protocol or in cables, since you mentioned 'custom-made DACs'. In order to understand what issue has been resolved in 2609 for 5940 concerning aggregated links, could you send me the link to that document?

TBH there is one thing that 'bothers' me in the network topology you are trying to build. You say "I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches." Those 5900AF, are they in IRF?
Another thing - "Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch". Even if 5900AF switches are in IRF (which means they both look like single physical device to neighboring hosts), what about those two Dell switches? Do they form a stack? If they don't, do they have any kind of MLAG to present themselves to the 5900AF IRF as one single LACP peer? For now it looks like two Dell switches work independently and present themselves with different SystemIDs to 5900 stack, therefore LACP process thinks it's a wiring error and doesn't allow second port to join the BAGG. In order this setup to work you need to make both Dell switches to present themselves with one common LACP SystemID to our stack, Dell calls this technology 'MLAG'.

It will help if you post an output of 'display link-aggregation verbose' in particular its section concerning that non-working BAGG. Then I will tell you for sure if my assumption about lack of MLAG on Dell's side is correct.

Ivan,

Thanks for the info. I feel I need to tell you "why" I was asking about the upgrade info. 

We are in the middle of deploying a new SAN at my organization. Part of the SAN includes (2) Dell S5212F0 switches. I was trying to build a bridged-aggregate interface using two ports on each of my two 5900 AF switches. Each of the two ports in the Bridge Aggregate interface would connect to one of the ports on each Dell switch.using custom made DAC cables. When we connect the the core using the bridged aggregate ports to the two Dell switches one of the two ports show "Current state: UP" and "Line protocol state: UP". The other port in the Bridged-Agg says "Line protocol state: UP" and "Line protocol state: Down(LAGG)". I was trying my best to fix this by redoing both interfaces, redoing the bridged agg, etc.etc. No matter what I do I cannot make this other port show up as "Line protocol state: UP".

I am sorry I did give you any context to my request. We had found an HPE support article for a "HP5940 AF" that said to upgrade to R2609 or later to fix this issue. Do you have any ideas on the first thing I should have asked :)

Thanks