Here we have the PGCon 2020 playlist:
My apologies for it taking so long.
Does anyone have Mailman 3 running on #FreeBSD?
If so, which zope version are you using? Did you install the mail/mailman3 port?
This post is NSFW. You've been warned. Ugly stuff here.
Today I was debugging a website issue after adding http://devgit.freshsource.org to webserver at home. The website was working via http://devgit-freshsource.int.unixathome.org but not via the public hostname. After figuring out why, I discovered two blog posts are needed.
The first is about split-dns.
Split-dns, according to some, should never be done.
I'm doing it. On a per-host basis, not for the whole zone.
Yes, http://devgit.freshsource.org is a single DNS zone file at home.
I am trying to locate an Eaton 5PX pedestal / stand.
None for sale on eBay. Do you have one in your junk pile which is no longer required? Would you let me buy it from you?
What I'm after is the stand like shown in this photo.
They are supplied with the UPS, and not sold separately.
I think most people toss them if they are using them in a rack.
As asked elsewhere, will I be soldering circuit boards?
No, just the wires, no boards or components.
I've been given an R720 and plenty of RAM (different donations, same source). The plan: get drive cages and drives, configure it as a FreshPorts node, put it back into the data center. You may recall I brought home a broken server back in late August. FreshPorts has been running on the old server since then.
I hope to crowd-source the funds for that new hardware.
I'm also working on getting FreshPorts onto AWS as well.
Hardware starts arriving today.
Next: flash the H710 mini into IT mode.
'zfs hold' avoids unintended 'zfs destroy' of a snapshot
From man zfs:
zfs hold [-r] tag snapshot - Adds a single reference, named with the tag argument, to the specified snapshot or snapshots. Each snapshot has its own tag namespace, and tags must be unique within that space. If a hold exists on a snapshot, attempts to destroy that snapshot by using the "zfs destroy" command returns EBUSY.
I ran tests over the weekend to make sure the rack shuts down smoothly when a power failure occurs. Just as important, I wanted to make sure it starts up properly and unattended. Usually I am not near the rack when power is restored.
It takes about 3.5-4 minutes to shut down.
From power restoration, it is about 7 minutes before all hosts are up, about 15 minutes for all jails to be running, and 25 minutes before Nagios is all green.
I testing the shutdown process of my UPS when using nut. I did this without turning off the UPS or letting the batteries run down.
10-Sep-2020 11:56:48.476 #7 0x800a7a736 in ??
10-Sep-2020 11:56:48.476 exiting (due to assertion failure)
Hmm, named on one host just crashed.
10-Sep-2020 11:56:48.475 resolver.c:5125: INSIST(dns_name_issubdomain(&fctx->name, &fctx->domain)) failed, back trace
10-Sep-2020 11:56:48.476 #0 0x2cfdf0 in ??
10-Sep-2020 11:56:48.476 #1 0x4b3e9a in ??
10-Sep-2020 11:56:48.476 #2 0x4005fb in ??
10-Sep-2020 11:56:48.476 #3 0x4065b9 in ??
10-Sep-2020 11:56:48.476 #4 0x40e000 in ??
10-Sep-2020 11:56:48.476 #5 0x40baf4 in ??
10-Sep-2020 11:56:48.476 #6 0x4d9a0c in ??
I've been using @Tarsnap since at least 2015-10-09
I recently decided to purge some old backups. Along the way, I found out just how much compression and deduplication are saving on storage space.