[Pdns-users] Issue with lightningstream replication not working

[Nico Cartron]

Answering to myself after I've been pointed out off-list that I was missing

lmdb-shards=1

in my pdns.conf

I don't know how I missed that, but as soon as I added that line, 
lightningstream sync worked straight away and I can now see all changes 
propagated in both directions - new zones, editing the content of a zone etc

Thank you Konrad! :-)

Cheers,

-- 
Nico

On 5/17/23 22:24, Nico Cartron via Pdns-users wrote:
> Hi,
>
> I'm testing PDNS Auth 4.8-beta1 with lightningstream on FreeBSD 13.
>
> I compiled PDNS manually with LMDB module (and gmake, as suggested in 
> the README), and am using Backblaze B2 as S3 backend.
> The Lightningstream log indicate that replication happens in both 
> directions, and I can indeed see the snapshots in my S3 bucket.
>
> But when I create a DNS zone on one PDNS server (with pdnsutil) and 
> populate it, I can't see that zone on the second PDNS server - a 
> pdnsutil list-all-zones doesn't show it.
>
> However, when I try to create that same zone on the second PDNS 
> server, pdnsutil tells me that the zone already exists!
> And surely enough, when I delete that zone on the first PDNS, then I 
> can create it on the second one - which shows that the 
> LMDB/Lightningstream workflow works.
>
> The Lightningstream status webpage (http://:8500) also shows the same 
> metrics for both instances.
>
> I must be missing something, but I'm having a hard time figuring out 
> what.
>
> I looked at the Lightningstream doc and everything looks good.
>
> My pdns.conf:
>
> local-address=192.168.x.y
> local-port=53
> launch=lmdb
> lmdb-filename=/var/spool/pdns-4.8/pdns.lmdb
> lmdb-random-ids=yes
> lmdb-flag-deleted=yes
> lmdb-map-size=1000
> lmdb-sync-mode=sync
> zone-cache-refresh-interval=0
> zone-metadata-cache-ttl=0
>
>
> My Lightningstream YAML conf file:
>
> instance: pdns
> lmdbs:
>   main:
>     # Auth 'lmdb-filename'
>     path: /var/spool/pdns-4.8/pdns.lmdb
>     schema_tracks_changes: true
>     options:
>       no_subdir: true
>       create: true      # optional for 'main', as auth will create it 
> on startup, if needed
>       map_size: 1000MB  # for create=true, make sure to match auth's 
> lmdb-map-size
>   shard:
>     # Auth 'lmdb-filename' plus '-0' for the first shard
>     path: /var/spool/pdns-4.8/pdns.lmdb-0
>     schema_tracks_changes: true
>     options:
>       no_subdir: true
>       create: true      # strongly recommended for shards
>       map_size: 1000MB  # for create=true, make sure to match auth's 
> lmdb-map-size
>
> storage:
>   type: s3
>   options:
>     access_key: XX
>     secret_key: YY
>     bucket: pdns
>     create_bucket: false
>     endpoint_url: https://s3.us-west-000.backblazeb2.com
>
> http:
>   address: ":8500"  # for status and metrics
>
>
> Cheers,
>