20227: Change default BalacePeriod to 6 hours.

This matches BalanceTimeout.

The previous time of 10 minutes was much to short, since sweeps are
not incremental, a full sweep on a moderately sized cluster could
easily take 8 minutes out of the 10 minute interval meaning
keep-balance would be running almost constantly.

Add upgrading note about keep-balance.

refs #20227

Arvados-DCO-1.1-Signed-off-by: Peter Amstutz <peter.amstutz@curii.com>

diff --git a/doc/admin/upgrading.html.textile.liquid b/doc/admin/upgrading.html.textile.liquid
index e48a27077df35fc989a413b8403b00b2c7dc2b90..4b45142a75df9792bef45e5d5de8e71fa19a7b0b 100644 (file)
--- a/doc/admin/upgrading.html.textile.liquid
+++ b/doc/admin/upgrading.html.textile.liquid
@@ -32,6 +32,10 @@ h2(#main). development main (as of 2023-03-06)
 
 "previous: Upgrading to 2.5.0":#v2_5_0
 
+h3. Slow migration on upgrade
+
+Important!  This upgrade includes a database schema update changing the integer id column in each table from 32-bit to 64-bit.  Because it touches every row in the table, on moderate to large sized installations this may be very slow (on the order of hours). Plan for the arvados-api-server package upgrade to take longer than usual.
+
 h3. Default request concurrency, new limit on log requests
 
 The configuration value @API.MaxConcurrentRequests@ (the number of concurrent requests that will be accepted by a single instance of arvados-controller) now has a default value of 64, instead of being unlimited.
@@ -48,9 +52,9 @@ There is a new configuration entry @CloudVMs.MaxInstances@ (default 64) that lim
 
 Using the obsolete configuration entry @MaxCloudVMs@, which was previously accepted in config files but not obeyed, will now result in a deprecation warning.
 
-h3. Slow migration on upgrade
+h3. Default frequency for running keep-balance has changed
 
-This upgrade includes a database schema update (changing an integer column in each table from 32-bit to 64-bit) that may be slow on a large installation. Expect the arvados-api-server package upgrade to take longer than usual.
+The frequency that @keep-balance@ will run (@Collections.BalancePeriod@) has been changed from every 10 minutes to every 6 hours.
 
 h2(#v2_5_0). v2.5.0 (2022-12-22)
 

diff --git a/lib/config/config.default.yml b/lib/config/config.default.yml
index 20d071680ee7af9144d63b442e5533886e931465..1919d7b704af0ce7c0d92fe4d0320229f84d0c51 100644 (file)
--- a/lib/config/config.default.yml
+++ b/lib/config/config.default.yml
@@ -534,7 +534,7 @@ Clusters:
       #
       # If SIGUSR1 is received during an idle period between operations,
       # the next operation will start immediately.
-      BalancePeriod: 10m
+      BalancePeriod: 6h
 
       # Limits the number of collections retrieved by keep-balance per
       # API transaction. If this is zero, page size is