Tape disappearing/reappearing in fabric

So I came into work and fell into a problem with backups. I noticed that my jobs have been running for 2 days and haven’t finished yet so I figured something was wrong. Checking the logs, I saw that 

fctl: [ID 517869 kern.warning] WARNING: fp(2)::GPN_ID for D_ID=20200 failed

 fctl: [ID 517869 kern.warning] WARNING: fp(2)::N_x Port with D_ID=20200, PWWN=500104 disappeared from fabric

fctl: [ID 517869 kern.warning] WARNING: fp(2)::N_x Port with D_ID=20200, PWWN=500104 reappeared in fabric

 scsi: [ID 243001 kern.warning] WARNING: /pci@3e,0/pci1166,142@10/pci1077,137@0/fp@0,0


REPORT LUN to D_ID=0x20200 lun=0x0 failed: sense key=Hardware Error, ASC=44, AS

CQ=0. Giving up

scsi: [ID 243001 kern.info] /pci@3e,0/pci1166,142@10/pci1077,137@0/fp@0,0 (fcp2):

 offlining lun=0 (trace=0), target=20200 (trace=914101)

genunix: [ID 408114 kern.info] /pci@3e,0/pci1166,142@10/pci1077,137@0/fp@0,0/medium-changer@w500104,0 (sgen5) offline

After seeing those messages I pulled up the handy cfgadm -la and got:

c4::500104           unavailable  connected    unconfigured failed

c4::5001041           tape         connected    configured   unknown

c4::5001042           tape         connected    configured   unknown

Which led me to try:

root@11:52:02 / 

$ cfgadm -c configure c4::500104

cfgadm: Insufficient condition

root@ 11:54:51 / 

$ cfgadm -f -c configure c4::500104

cfgadm: Library error: report LUNs failed: 500104

So what to do next? I checked some other servers that are sharing the same tape device and they all had the same problem so I thought it must be something with that. I went back to my windows training and rebooted the SL500 (using the SLConsole from Sun). Once the reboot was done, I got a new output showing:

c4::500104f000           med-changer  connected    unconfigured unknown

Do a quick little command and everything was good again:
cfgadm -c configure c4::500104f0007ee1bf
This entry was posted in hardware, software and tagged , , . Bookmark the permalink.

8 Responses to Tape disappearing/reappearing in fabric

  1. Christian says:

    experiencing the same problem, only with our SAN instead of a tape drive.
    Everything looks the same up until “So what to do next?”.

    My problem is that I can’t really reboot the SAN.

    This issue came after we replaced one of the fiber switches. Only Solaris server have the problem, fortunatly there are 2 path’s to each disk and one path works. I tried rebooting one of the servers that experience this issue and after a reboot everything works fine but I can’t reboot the other 5 servers, it’s production systems that hold well over 20 databases and application servers.

    Our Windows-servers and VMware clusters work fine after replacing the switch, only Solaris server that have problems.. sigh 🙁

  2. Christian says:

    Today when the hardware ppl changed the corresponding FC-switch at our other server location the path’s began to work again. Wonderful and horrible at the same time. I hate to not understand why things went wrong.

  3. Ryan says:

    I do hate it too when I don’t understand why something has happened. Glad everything is working for you now.

  4. mahesh kumar says:

    I am allso facing same issue not only master server as well as media servers also getting tape drive fluctuation issue…kindly suggest.


  5. Ryan says:

    Hi Mahesh,

    I had something similar happened and it was a bad sfp in our fc switch that our tape drive was plugged into. You should try logging into your fc switch and see if you have a lot of errors on that specific port to your tape drive. If you do, you can reset the error count and wait to see if more occur, or try plugging your tape drive into another sfp on the switch.

  6. mahesh kumar says:

    But fluctuation issue is not for one drive but all Library as well as Medium changer. We have changed FC cable also but no result. please suggest if you know…..

  7. Flex says:

    What was the resolution for the issue…I am experiencing same issue except only when i turn the tape drives up to 8GB… If they are at 4 or 6 GB things appear to operate fine. The Fiber cards on the V440 are 2GB max
    fctl: [ID 517869 kern.warning] WARNING: fp(1)::GPN_ID for D_ID=ee1500 failed
    fctl: [ID 517869 kern.warning] WARNING: fp(1)::N_x Port with D_ID=ee1500, PWWN=500308c38c19e009 disappeared from fabric

  8. Ryan says:

    Hi Mahesh, I’m sorry I didn’t have anything to offer but I’m guessing after a year you got it solved? Probably through warranty support?

    @Flex, I haven’t ever seen that problem before since I’ve only used 4Gb switches and 2Gb and 4Gb hbas. I wonder if there something that is not backwards compatible with 8Gb to 2Gb speeds. You may want to open a service request with Oracle or the fiber card vendor.

Leave a Reply

Your email address will not be published. Required fields are marked *