UPDATE - 4:15pm

Anchor migration is more-or-less complete for now (at least, "big stuff" is done). Known details at this point,

  • The first time you connect via SSH, you may be told the host ID has changed. This is normal. Please accept the change or delete old host ID info if required and you can connect normally with your normal username and password
  • NFS mounts to all client systems appears to be working properly now (Solaris, Linux, etc)
  • I've had one sunray user who can't login / will require more debugging, so if you have problems - please let me know. Most of the time simple/quick fixes are all that is required.
  • Note that your home directory may have "more stuff" in it than it did at the start of the day: This is due to "very conservative" file transfer options, coupled with the "first transfer" of anchor data having taken place over 1 month ago. Files which you may have moved since that time - may suddenly be present in the "old" location (with older time stamp) in addition to the new location. Hopefully this won't cause significant confusion or problems for anyone.

Please let me know if any problems arise. Email is the best way to get things "into the queue".

Thanks

--Tim Chipman


UPDATE - 2:45pm PLEASE NOTE:

The deployment of the new anchor server is well underway. At this point,

  • the old anchor has been reconfigured to a new identity
  • the new anchor has been reconfigured as "Anchor"
  • all user data has been cleanly synchronzied to the new anchor
  • the new sunray server is up and running

KNOWN PROBLEMS AT THIS TIME:

  • If you ssh to "anchor" you may be presented with an error message indicating that the host ID

has changed. This is due to the migration process (the host HAS changed, but intentionally). To gain access, you may need to delete a key file as directed by the error message.

  • NFS mounts from anchor to solaris systems (such as jade, kt-4800) are NOT working presently.

This will be debugged thisafternoon

(In contrast, NFS mounts to linux servers - such as mahone and hartling - are working fine)

I'll be debugging things steadly and will provide an update later in the day.

--Tim Chipman


OLDER POSTS FOR REFERENCE:


POSTED MONDAY AM -- PLEASE BE ADVISED:

Final prep work for the migration of anchor is proceeding smoothly, despite some issues with the power outage over the weekend.

I believe we're on-track for anchor to go offline at 12-noon today, as per the plan outlined last week.

Please save your work and log out of anchor starting at 11:45 and no later than 11:55.

Thanks,

Tim Chipman