diff options
Diffstat (limited to '')
| -rw-r--r-- | 2024/organizers-notebook/index.org | 28 |
1 files changed, 14 insertions, 14 deletions
diff --git a/2024/organizers-notebook/index.org b/2024/organizers-notebook/index.org index bbb1032a..40241c26 100644 --- a/2024/organizers-notebook/index.org +++ b/2024/organizers-notebook/index.org @@ -213,7 +213,6 @@ CLOSED: [2024-11-02 Sat 11:38] links ; <= 10:30 regex ; <= 12:00 sun or >= 12:00 sat learning ; <= 13:00 - language ; <= 12:00 (org-teach :buffer 15) ; any (hyperbole :buffer 15) (lunch :start "12:00") @@ -474,7 +473,7 @@ Some convenient commands are defined in emacsconf-el:emacsconf-erc.el. :PROPERTIES: :CUSTOM_ID: decisions :END: -** BigBlueButton replacement +** BigBlueButton replacement (BBB) :PROPERTIES: :CUSTOM_ID: bbb :END: @@ -491,15 +490,16 @@ Recommendation: We spin up a shared CPU 4GB under bandali's Linode account (for bandwidth pooling and easier -reimbursement), do the setup/testing over the next -few weeks, scale up to dedicated CPU 16GB a day or -two before the conference, and keep it at that -level until the recordings are all done. I have -access to that Linode account, so I can set it up. -If bandali can configure emacsverse.org to be -handled by Linode, I can manage the DNS changes, -or he can change bbb.emacsverse.org to the IP -address of the new node. +reimbursement), do the setup/testing, scale down +as small as possible over the next few weeks, +scale up to dedicated CPU 16GB a day or two before +the conference, and keep it at that level until +the recordings are all done. I have access to that +Linode account, so I can set it up. If bandali can +configure emacsverse.org to be handled by Linode, +I can manage the DNS changes, or he can change +bbb.emacsverse.org to the IP address of the new +node. Some options: @@ -518,15 +518,15 @@ Some options: - (* 31 0.125) 3.875, oh, maybe we can get away with 10GB storage and add another chunk of 10GB midway if we need it, or extend the dedicated - Actually, block storage probably not needed if we're going to keep the big instance up until the recordings are done - Ideas + - *We're here -- Option B: Run a slightly larger shared CPU instance from now until shortly before the conference, then scale up* in case BBB does not work for installation/testing with less memory; block storage probably not needed + - (* 0.07 24 7 3) USD 35.28 for 8GB, total (+ 35.28 36.96) USD 72.24 + - (* 0.14 24 7 3) USD 70.56 for 16GB, total (+ 70.56 36.96) USD 107.52 - *Option A: Run a small instance continuously, scale up for the conference, pull the recordings off, spin the instance down* - est total USD 60+tax - There are three weeks between now and the conference (* 0.04 24 7 3) USD 20 - Last year, recordings were available within three days after the conference - Let's plan for a week at high capacity. (* 0.22 24 7) USD 36.96 - and a month of 10GB, possibly up to 30GB. USD 3 - That's probably small enough that it doesn't make as much sense to try to fully spin down for a week in between now and the conference (* 0.04 24 7) - which would save USD 6.72 but require figuring out backing up, etc. - - *Option B: Run a slightly larger shared CPU instance from now until shortly before the conference, then scale up* in case BBB does not work for installation/testing with less memory; block storage probably not needed - - (* 0.07 24 7 3) USD 35.28 for 8GB, total (+ 35.28 36.96) USD 72.24 - - (* 0.14 24 7 3) USD 70.56 for 16GB, total (+ 70.56 36.96) USD 107.52 - Ask FSF if we can borrow their BBB or Galene, maybe limit it to cycling among 6 or 7 rooms - Waiting for feedback; Corwin has asked them |