fabric1-core-e07-192168162022:root> switchshow
3 3 id N4 Online Loopback->Port 3
有一台EVA6400的存储，两个控制器都坏了。
8个端口所连接的FCSW上面都显示端口状态Loopback->Port ...
非常奇怪，在网上也找到一篇类似的文章:
Hello,
last week when I get connected via telnet to one of our switches (silkworm4900) and when I executed a switchshow command, two of the 64 ports look like this:
15 15 id N1 Online Loopback->Port 15
40 40 id N1 Online Loopback->Port 40
When I checked out what were in those ports I found that both of them belong to the 2 ports of one of our IBM 3592 Tape drive. Then I checked too in our last SAN configuration with SanHealth the section "fabric summary" and there those ports looked like this for both ports:
speed 1Gbps, status online, type false, Avg Perf 0MB/s, Max Perf 0MB/s, other fields were empty.
Then I ran the command portdisable just for port 40 and then I enabled it but after portenable the port kept in state in_sync, i disabled it again and i went to the data center and i unplugged the port 40 and plugged it again but in a diferent port but I didn?t get any answer, there the port kept in_sync too, and it was the same with port 40.
15 15 id N4 In_Sync
40 40 id N4 In_Sync
Now we are waiting for a test of the fibre cable and for the tape unit, but in the midtime I really want to understand what could cause this situation or why without any change in port type configuration in the switch those ports got in this state and when i disabled them they change from online to in_sync state. What is the meanin for loopback->port and what situations can cause the change for the porttype
回复:
Hi Andre
Here’s some info that may help....
Loopback...
You can put a port into loopback mode for testing and diagnostics, this is usually done with a loopback cable connected.
However some commands just test the internals of the ASIC in loopback, commands such as crossporttest.
But a port must be manually placed into loopback mode by someone???
In_Sync Port status....
Definitely check the device and the optics that does sound like the most likely cause.
If you disable/enable plug out/in a port – the device on that port goes through a whole handshake process and a Fabric login (FLOGI) and Port Login (PLOGI). Then end result is that the port gets an on-line status and the device gets registered in the name server. In the “old” days device drivers sometimes mishandled the FLOGI/PLOGI timing and we used to see the occasional issue with ports coming online. But I haven’t seen this in at least 6 years. The switch tracks this login/port change stuff in the portlogdump and your support provider should be able to decode the portlogdump to see exactly what is going on when that device is plugged in. That said, if it’s an optics problem, then you will get nothing in the portlogdump as the switch is not seeing anything.
不知道是不是EVA6400控制器坏掉之后会导致光纤模块自动回环呢？
3 3 id N4 Online Loopback->Port 3
有一台EVA6400的存储，两个控制器都坏了。
8个端口所连接的FCSW上面都显示端口状态Loopback->Port ...
非常奇怪，在网上也找到一篇类似的文章:
Hello,
last week when I get connected via telnet to one of our switches (silkworm4900) and when I executed a switchshow command, two of the 64 ports look like this:
15 15 id N1 Online Loopback->Port 15
40 40 id N1 Online Loopback->Port 40
When I checked out what were in those ports I found that both of them belong to the 2 ports of one of our IBM 3592 Tape drive. Then I checked too in our last SAN configuration with SanHealth the section "fabric summary" and there those ports looked like this for both ports:
speed 1Gbps, status online, type false, Avg Perf 0MB/s, Max Perf 0MB/s, other fields were empty.
Then I ran the command portdisable just for port 40 and then I enabled it but after portenable the port kept in state in_sync, i disabled it again and i went to the data center and i unplugged the port 40 and plugged it again but in a diferent port but I didn?t get any answer, there the port kept in_sync too, and it was the same with port 40.
15 15 id N4 In_Sync
40 40 id N4 In_Sync
Now we are waiting for a test of the fibre cable and for the tape unit, but in the midtime I really want to understand what could cause this situation or why without any change in port type configuration in the switch those ports got in this state and when i disabled them they change from online to in_sync state. What is the meanin for loopback->port and what situations can cause the change for the porttype
回复:
Hi Andre
Here’s some info that may help....
Loopback...
You can put a port into loopback mode for testing and diagnostics, this is usually done with a loopback cable connected.
However some commands just test the internals of the ASIC in loopback, commands such as crossporttest.
But a port must be manually placed into loopback mode by someone???
In_Sync Port status....
Definitely check the device and the optics that does sound like the most likely cause.
If you disable/enable plug out/in a port – the device on that port goes through a whole handshake process and a Fabric login (FLOGI) and Port Login (PLOGI). Then end result is that the port gets an on-line status and the device gets registered in the name server. In the “old” days device drivers sometimes mishandled the FLOGI/PLOGI timing and we used to see the occasional issue with ports coming online. But I haven’t seen this in at least 6 years. The switch tracks this login/port change stuff in the portlogdump and your support provider should be able to decode the portlogdump to see exactly what is going on when that device is plugged in. That said, if it’s an optics problem, then you will get nothing in the portlogdump as the switch is not seeing anything.
不知道是不是EVA6400控制器坏掉之后会导致光纤模块自动回环呢？
5315
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
