- Got processed at the Hill Ave. office.
- Met everyone. Got the grand tour. Started learning the configurations
of the computers.
- Prepared a logical map of the Ethernet network wiring in the two houses.
Also went into the steam tunnels to trace the network wiring across the
street.
- Went to CCO to learn the procedure for obtaining software for the VMS
machines.
- Installed FORTRAN and C compilers on the Alpha workstation. Also
installed GKS on the Alpha.
- Added LAT to the system startup files on the Alpha so that I could get to
it from my terminal.
- Sat in on the USGS staff meeting to learn more about how things work
around here.
- Sat in on a C++ programming class.
- Got the Ethermon package to do traffic monitoring on the Ethernet.
- Began implementing the Diskeeper disk defragmenter on some of the disks.
- Made temporary configuration changes on two of the workstations as an
experiment in system tuning. The users reported improved performance.
After the disk defragmenter has been fully implemented, I will make these
configuration changes permanent.
- Discovered so possible file structure problems on one of the disks. I
have scheduled some time on Monday morning to take that disk off line for
maintenance.
- Went to the USGS weekly staff meeting.
- Performed regular system administration tasks. Created a couple of
accounts, and disabled a couple of old accounts that were no longer being
used.
- Moved the VAXstation 4000-90 from Bob's office to my office. I had to
fix a problem with the TCP/IP configuration on it before it would work
correctly.
- Met with the representative from Digital to begin preparing a quote for
buying another Alpha workstation.
- Worked with Bob to clear up a problem with two of the VAXcluster systems
on Wednesday. Made some minor configuration changes on one of the
workstations at the same time.
- Continued working with the Diskeeper disk defragmenter. I have set it
to run nightly from 9PM to 5AM.
- Experimented with a 10base-T Ethernet hub that we have here. It turned
out to have some problem in its transceiver, so it kills the network when
it is plugged in.
- Discovered that the system of doing disk backups is very ad-hoc, and is
in need of scheduling. Met with Bob, Kate, and Riley to work out the
details of a set backup schedule.
- Fixed a problem with batch jobs not running correctly. It turned out to
be caused by glitch that left one of the disks mounted write-locked on the
main VAX.
- Fixed a problem with the QMS postscript print queue. The queue
definition was incorrect, and the queue would not print. I deleted the
queue and redefined it and it works properly now.
- Reconfigured the system disk on BIGONE to recover some disk space by
juggling and adjusting page, swap, and dump files. This required an
early-morning shutdown and restart of BIGONE, MECCA, MALIBU, and AVALON.
The restart was completed by 8:00 AM.
- Moved the page and swap files for node INDIO to the local disk. This has
helped performance on that node, as well as freeing 65000 blocks on the
BIGONE system disk, and cutting down on LAVC traffic on the Ethernet.
- Continued working with the Diskeeper disk defragmenter. I have set it
to run nightly from 9PM to 5AM on the system disk, and have changed the
user disk job to run weekly on Thursday nights.
- Met with Riley and worked out a preliminary backup schedule. This would
back up important data three times a week, and each of the 22 disks would
be backed up at least every other week. This has been submitted to Bob,
Doug, and Kate for comment.
- Wrote a command procedure to check all nodes of the cluster to see that
all of the disks are properly mounted to avoid problems like the batch job
problem that occurred last week.
- Wrote a command procedure to create new ERRLOG.SYS files for each node in
the cluster. This had not been done for a couple of years, and the error
log files had grown quite large, which made checking on recent device
errors very slow. Now it can be done automatically on a regular basis.
- Did some housecleaning in the system user authorization file. Wrote a
program to check all accounts and flags accounts that had not had any
activity in the past year. After consulting with Bob, Doug, and Kate, I
disabled about 10 unused accounts.
- Began writing a program to automatically generate the DIS$USERS
distribution list directly from the system authorization file. I
discovered that the existing list has not been updated in quite some time.
- Discovered that several of the disks on the system are not owned by the
[SYSTEM] account, as they should be. This is not a critical problem, but
should be addressed at some point.