Before the "Big Reorg" - May 12, 2002

Our server closet was in disarray. We had to get rid of some flakey hardware and rack up some new stuff in as little time as possible. Here are some shots from the day before the "Big Reorg."

Top of the table - On the left are the four DLT drives attached to our science and user database servers, (for database/network backups and splitting of data tapes into workunits). On top of those is a good ol' Wyse terminal connected to all the important machines (Note: there's not enough resolution to read the text on the screen, so don't bother trying). In the middle are two of the eight six-packs of external disks attached to the science server. On top of those is a switch (from which most of the machines in the room get their network connectivity), and the a/b/c/d box for the terminal. On the right is a Sun Ultra 60, currently both our web server and SETI development file server.
Below the table - three of our main servers, all Sun E450s. Starting from the left we see the user and science database servers (sitting on top of those two are four of the eight six-packs of external disks attached to science database), and finally the data distribution server. On top of the data server is an old, used Sun Ultra 2 which is our single in-house ftp server from which users download client software.
Obscured by the front wall of the closet, what you see here is our NetApp 3-Terabyte file server. It doesn't look very exciting from this angle, but it is. We keep a lot of stuff on it - all the work units to be distributed to users, our science database, scads of SERENDIP data products, entire disk images of DLT tapes, among other things. It's thanks to this machine that we'll finally be able to lose those eight six-packs which seem to have multiplied like tribbles. That's the data server in the lower right corner.
A different, more personal view of our data (and ftp) server. That single machine belches out data at the average rate of 25-30 megabits/sec, which accounts for about 1/3 of the total bandwidth of the Berkeley campus. That's more bandwidth than the entire internet about 7-8 years ago. Or so we've been told. Anyway, there's a bunch of stats like this about sagan which seem to impress a certain subset of the population.
A small sample of the nest of wires behind all the E450s. What a mess! All kinds of wires are represented here: SCSI, serial, ethernet, power, and fiber. This is but one of the reasons we're reorganizing the closet. Those are two of several UPSs on the right.
Hiding back behind the NetApp rack is this pile of stuff. The big thing on the bottom is a Sun E3500 acting as the master database server, i.e. the database that contains cleaned-up, reduced results from the science database. Above that, in order, are two Sun A1000 disk storage racks for this database, then an old Sparc 20 which is an administrative machine and SERENDIP data slurper, then another DLT drive, and an 8mm tape drive.
And here's a nice capture (provided by danw) of the whole closet from the front. A thing of beauty. Includes the Flag of Earth, the air conditioner, and a peek at the Compaq disk arrays sitting on the floor behind the E450s (attached to the science database for extra scratch space).
And there's a very old Sun IPX, which was one of the first Sun IP machines here in the space lab. In a past life it was considered a powerful computing tool. It retired into service as an account server and extremely slow desktop machine. On the very day SETI@home launched in 1999, our ftp server was overwhelmed by the unexpected high demand, so Jeff and Matt franctically converted this machine into an extra ftp server. They even cannibalized eight 1-Mbyte SIMMs from a second IPX to put in this machine and enable it to handle about 10-20 more ftp requests per minute. Those were the days. Now it's barely holding its own as a doorstop.
Now, hanging out in our new room 329, is this handsome rack stuffed with new Sun equipment. Starting from the top are two netras, and then two E420Rs. Disappearing below the edge at the bottom is a D1000 drive array, two more E420Rs, and finally one more D1000. These machines haven't been assigned hard tasks yet, but are currently acting as (very useful) extra CPUs and failover machines in case one of the E450s in the closet kicks the bucket.
And here's what will eventually become the second generation SETI data recorder. It's an LH6000R HP server running linux. Pretty powerful machine. Below the keyboard is a 9-tape DLT "stacker." You don't actually stack the tapes as much as you wedge them into various corners of its big ol' mouth.
On the surface, this is pretty much a dull photo. But what you are seeing here is room 329. Currently we all squeeze into one lab (room 325) and have to share it with very noisy computer equipment and piles of paperwork and spare parts that defy organized storage. So having this extra room two doors down being availed to us is a rather important moment for SETI@home. We might actually have space for students to work here over the summer! Plus this'll be some handy-dandy buffer space when we re-organize the server closet. Obviously we haven't done much with it yet, but we will.

Album Menu  -  Next (Later) Album

©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.