Rebooted the server to upgrade ZFS to git master to test the newest ZFS feature: sequential scrubbing. Appears to save 15 minutes of scrubbing time. Server upgrade still coming soon.
Cogeco has increased our network speed to 80/15. Still holding out that they bring symmetrical gigabit fiber to Hamilton soon.
Brought the router down for a couple hours this morning to replace the non-functional wifi card. All is well.
I have upgraded our aging router/firewall to brand new hardware, and should serve us well for many years to come:
Core i7 7700 @ 3.6GHz
8GB Single Channel DDR4 2400MHz
2x 250GB Samsung 850 EVO SSD, in ZFS mirror
Realtek RTL8111G Gigabit Ethernet for the WAN port
2x Intel 82574L Gigabit Ethernet for the LAN + DMZ
Realtek RTL8192CE G/N Wireless (will be replacing this soon with an Atheros-based card as this card appears to be defective)
I will be upgrading the server in the coming weeks to the same processor, with 16GB of memory, which may sound like a step backwards, but we weren’t really using that extra memory for much besides disk caching. I may decide to upgrade to 32GB anyways. Stay tuned.
After 2 months of attempts, we have successfully upgraded to 4.4(hardened) and the latest ZFS development branch. The latest ZFS development branch contains a new feature called ABD which is meant to reduce memory waste by changing from the standard slab allocator for cache objects to a scatter list. Initial testing on other machines showed nice improvements. Attempted to upgrade to 4.8(hardened) at the same time, which ran into some roadblocks with boot-time kernel panics. Will try upgrading to a newer kernel in the future.
UPS batteries had failed during a power outage. Batteries have been replaced and all is well.
Been a wacky month.
Had to shut down for an hour on April 25th, due to extended power outage.
The past 2 weeks have been spent trying to upgrade the storage on bloomfield. This work has been completed, see the shiny new specs on the status page.
Will have to power down briefly in the next week or so to replace the UPS batteries. More updates to follow.
After a year since the last system outage, we started having issues with both our router and the shell server. Maintenance has been performed and is now complete. All services should be online. Sorry for the inconvenience.
We suffered a router failure on the night of Tuesday, October 14. A replacement was in the works, but the failure occurred at an unexpected time. The replacement has been put in place, but we’re still working out the bugs, so bear with us for a couple days. Thanks.
Recently moved, so the server was offline for 2 weeks. Server is back online, and been brought up-to-date. Cogeco has also upgraded our connection to 30Mbps/10Mbps.
If something looks broken, let me know so I can fix it.
Had to reboot the server for our yearly log rotate. Took about 40 minutes. Server back online.