VectorNTI network licenses purchased

We heard the clamor, and we've responded (within our budgetary constraints). We will be receiving (stay tuned for the date) a 10-simultaneous-user network license for VectorNTI. Though we wished Invitrogen had maintained their Macintosh version, dropped a number of years ago, and while we wish that VectorNTI had remained free to academics, their academic pricing-policies allowed for a reasonable license size.


Status update: RCBIG1 and RCBIG2

Both RCBIG1 and RCBIG2 have had half of their 250 GB drives replaced with Terabyte drives. At the same time, their operating systems were upgraded from Fedora FC3 to Fedora FC5. We're investigating if there would be any negative implications to moving to Fedora FC7, at which point we would be able to take advantage of regular, remote updates.

Next step on these machines is to move all files from the old drives to the new drives, and then replace the drives on the other half of the machines. There should be no downtime involved in this, except, briefly, when accounts are being moved.


RCBIG2 unavaialble Monday, 2/24 from 9:00 am to 10:00 am

The internal drive controller, which keeps a mirror of the system drive, on RCBIG2 reports that one of the drives is dead. In order to get at the internal drives, we have to take the machine down for an hour on Monday morning, 9:00 to 10:00.
Matt


Server Upgrade, Next Step

RCBIG1 has had one of its RAID controllers upgraded to 1Tb drives and is now up and running. The next target will be RCBIG2. The contents of the DATA1 partition will be moved to DATA2. Once that is completed, the drives that now make up DATA1 will be removed and replaced with Terabyte drives.
When the new RAID is rebuilt, the contents of DATA2 will be moved to the new RAID, after which, DATA2 will be replaced with a terabyte drive RAID. Whew.
The result, though, is that our old servers, which held about 4 Tb of RAID, will hold 20 Tb.


Working on RCBIG1 -- RCBIG1 Unavailable

[ monday, 2/16/09, 9:05 am ] RCBIG1 available, upgraded to FC5. Initializing RAID on Controller 2 allowed RCBIG1 to boot normally. Please let us know if you encounter any problems.

[ monday, 2/16/09, 12:15 am ] RCBIG1 wouldn't reboot. Initializing RAID5 on controller number 2. Upgrade attempted earlier yesterday failed on a bad CD. Don't yet know why it wouldn't boot; suspect it is because Controller 2 had /no/ RAID5. will attempt reboot and upgrade later today.


Size matters: upgrades to RCBIG, RCBIG1, and RCBIG2

We're in the process of upgrading RCBIG, RCBIG1, and RCBIG2.
At one point in time, these servers were BIG, but now their 250 GB drives are starting to look puny. We are replacing those drives with 1 Tb drives in as non-disruptive a manner as possible. The first stage is about to begin.


Difficulty Restarting Web Server on research.dfci.harvard.edu & fix

The webserver on research.dfci.harvard.edu appeared to be dead. Restarting produced the following error:

[Wed Jan 28 20:40:35 2009] [notice] LDAP: Built with OpenLDAP LDAP SDK

[Wed Jan 28 20:40:35 2009] [notice] LDAP: SSL support unavailable

[Wed Jan 28 20:40:35 2009] [crit] (28)No space left on device: mod_rewrite: could not create rewrite_log_lock
Configuration Failed


Syndicate content