07-30-2010 09:12 PM
We have a VIO Server connected to DCX Switch.. Portshow shows that 2 VIO Clients have logged in. And we see that Virtual WWPNs are logging into the Storage Frame DMX4. However VIO Client does not see the LUNs. Here is the Port Show output.
What is the significance of "invalid_word" count which is so high. I do see that Lr_in, Lr_out, Ols_in, Ols_out values are normal. So I think there is no SFP or cable problem. enc_out value is 2.7m. when I looked at porterrshow output.
Can someone assist in diagnosing this issue.
portFlags: 0x1024b03 PRESENT ACTIVE F_PORT G_PORT NPIV LOGICAL_ONLINE LOGIN NOELP LED ACCEPT
portState: 1 Online
portPhys: 6 In_Sync portScn: 32 F_Port
port generation number: 22752
state transition count: 122
portWwn of device(s) connected:
LE domain: 0
FC Fastwrite: OFF
Interrupts: 0 Link_failure: 1 Frjt: 0
Unknown: 17 Loss_of_sync: 6 Fbsy: 0
Lli: 94 Loss_of_sig: 2
Proc_rqrd: 13490 Protocol_err: 0
Timed_out: 0 Invalid_word: 2740339
Rx_flushed: 0 Invalid_crc: 0
Tx_unavail: 0 Delim_err: 0
Free_buffer: 0 Address_err: 0
Overrun: 0 Lr_in: 1
Suspended: 0 Lr_out: 1
Parity_err: 0 Ols_in: 1
2_parity_err: 0 Ols_out: 1
Port part of other ADs: No
07-31-2010 04:51 AM
have you done the zoning properly, have u saved it with cfgadd, cfgsave and cfg enable, you can use fcping, nscamshow whether the inititor or target ar talking ot each other.have you published the LUN in a properway .have you checked the WWNs of the HBAs and storage ports?
If everything is ok, doa portstatsclear on the HBA and storage port, then check through portstatsshow and porterrshow.
07-31-2010 10:27 AM
Zoning has been done well. For about 15 days, the VIO Clients were able to access LUNs through the Physical Server. After some time both VIO Clients sharing the Same Physical HBA are not seeing the LUNs.
I did run fcping between each VIO (Virtual WWPN) to Storage Port on the Frame (WWPN). It went well and no frames missing, no time out, no rejections.
I did clear the error count (portstatsclear ) on the Port which is connected to Physical HBA of the VIO Server.
I will monitor for a day more and will check with Unix Admin and let you know. Running the command, Portstatsclear, does it help in any way ?
12-14-2010 07:23 AM
Unless I've somehow missed it, I don't see an answer to the original question, "what is the significance of a very high invalid_word count?" I'm having the same problem, and would like to know the answer.
02-07-2011 07:00 AM
This is related to the wrong portcfgfillword on the SAN switch port which runs on 8Gbit.
I know that this is an old thread.
But maybe some one is intrested in this.