Allow switching from limits-based to interest-based retention in stream update #4361
Conversation
| // retention then we will need to check the ack floor of all consumers. | ||
| if ocfg.Retention != cfg.Retention && cfg.Retention == InterestPolicy { | ||
| toUpdate := make([]*consumer, 0, len(mset.consumers)) | ||
| for _, c := range mset.consumers { |
There was a problem hiding this comment.
One important piece, if the consumer is retention interest, then its replication factor must be same as stream, so need to adjust that here and add in an R1 consumer to the test above etc.
| c.retention = cfg.Retention | ||
| c.mu.Unlock() | ||
| if c.retention == InterestPolicy { | ||
| c.checkStateForInterestStream() |
There was a problem hiding this comment.
Nice! I saw this and was confused for a second then realized what you did here..
| t.Helper() | ||
| if bytes.Equal(a[:], b[:]) { | ||
| t.Fatalf("require not equal, but got: %v != %v", a, b) | ||
| if a == b { |
There was a problem hiding this comment.
Will this do the right thing for 2 different byte slices that have same contents?
There was a problem hiding this comment.
The function was previously taking a, b [32]byte and fixed-size arrays can be compared using ==, so it will continue to work for what it was being used for.
Trying to pass in slices instead of fixed-size arrays will result in a compile-time error as they are not comparable.
neilalexander
force-pushed
the
neil/switchtointerest
branch
2 times, most recently
from
fc4ea1c
to
8168aa2
Compare
neilalexander
force-pushed
the
neil/switchtointerest
branch
2 times, most recently
from
62537ec
to
d752c76
Compare
…am update Signed-off-by: Neil Twigg <neil@nats.io>
neilalexander
force-pushed
the
neil/switchtointerest
branch
from
d752c76
to
d7f76da
Compare
This should make it possible to switch from limits-based retention to interest-based retention on an existing stream.
Signed-off-by: Neil Twigg neil@nats.io