<html><center></html>STATUS updates<html></center></html>
Some links of interest:
I discovered that I did run a cable between the NFS's… tried bringing NFS1 back up, but it caused load to be extraordinarily high and impacted performance, so I took it back down. Will likely aim for a weekend resync.
Spent pretty much the whole day in the LAIR— 9am-6pm. It was a blast. The multiseat machines held up admirably.
My classes pretty much all start tomorrow.. taking advantage of this free Tuesday to put finishing touches on things. Claimed one of the pod spots by LAIReast as my unofficial office.
Spent some quality time on this Saturday applying some sanity to the network. Made some pretty diagrams.
The semester starts next week! We'll be ready. Deployed the interim replacement c107a-fs.
Yet more arranging, organizing.
Continue to assert some order to the new room.
May have gotten LAIReast back up/juicebox talking to the outside world…
Lab46 is once again accessible from the world… new IP: 143.66.50.9
Further refinements made. Some attempt at getting LAIReast back up- figuring out VLANs, running cables, etc.
The second of two big hauls took place, and we began our first re-assembly of the LAIR.
Got LAIRwest back up and running, no campus connection as of yet- scheduled for tomorrow.
The new pod layouts have commenced. Tri-Force design.
Attempting a second Fedora install on the second of 3 or 4 flake pods… both to verify my ability to reproduce similar results, but also to start contemplating documentation. Both drives so far have been 160GB SATAs, which seems a waste of resources… once the dust settles, I may well want to swap them out with smaller drives (or figure out how to boot from network and into RAM).
Looks like that didn't fix it… they persist.
Turns out the issue still persisted… added more options… the following stanza, as a whole, was added today to address the problem (in /etc/dovecot/dovecot.conf):
mmap_disable = yes dotlock_use_excl = no mail_nfs_storage = yes mail_nfs_index = yes lock_method=dotlock
I thing I may have finally solved the dovecot time sync issues… not sure why they were only present now, with only 1 NFS server present, but hey…
According to:
The option lock_method=dotlock might work well.. and sure enough, it did. There are some other NFS options, but they are not currently applied (as far as I can tell).
B&G arrived to start moving stuff… I spent the morning (and pretty much the day) wrangling equipment into piles to be taken to Goff Road.
Trying to address the persistent claimed time sync issues with dovecot… all involved machines have been manually rsync'ed against caprisun.. clocks do not appear to be a problem. Issue persists.
It has been decided we are to move the LAIR to Goff Road for the semester, room J106.
I ventured over there and took a look… a little smaller, but could conceivably work in the short (semester) term.
B&G will be assisting in our move… preparing necessary equipment for transport.
With mounting student and user demands for Lab46 availability to resume, I finangled some things to get the LAIRwest rack back up and running.
As anticipated, there were no problems standing in my way.
Logs are generally noisier due to various co-services in LAIReast and LAIRsouth being unavailable, and there's some interesting time sync error going on with dovecot on mail.
<html><center></html>
<html></center></html>