Jakub Sokołowski
74f980b956
status: disable connecting to bridge instance
...
Supposed to be a temproary fix for performance issues:
https://github.com/status-im/infra-status/issues/14
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-08-30 20:28:38 +02:00
Jakub Sokołowski
55c7b95239
status: fix certbot stopping nim-waku container
...
Otherwise it can't use ports 80 and 443.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-08-29 17:42:40 +02:00
Jakub Sokołowski
46238b40c3
status: enable native websocket port
...
https://github.com/status-im/infra-status/issues/13
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-08-22 19:27:45 +02:00
Jakub Sokołowski
64229615c1
status: lower retention to 14 days
...
Hosts are running out of disk space.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-08-11 19:39:55 +02:00
Jakub Sokołowski
96e0178d8a
use new format of specifying service for waku-peers
...
https://github.com/status-im/infra-role-waku-peers/commit/be28c829
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-07-26 21:21:28 +02:00
Jakub Sokołowski
335542a504
status: lowe SWAP file size to 2 GB
...
Default of 4 GB is too big.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-07-11 15:51:15 +02:00
Jakub Sokołowski
f903864fc7
waku-peers: bump timeout to 20 seconds and 5 retries
...
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-06-23 14:31:12 +02:00
Jakub Sokołowski
f695b346ed
waku-peers: use new variable names, add timer
...
Hourly timer avoids the need to have to re-run Ansible frequently.
https://github.com/status-im/infra-role-waku-peers/commit/895d6f07
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-06-22 13:22:10 +02:00
09b3d95ad6
fix: sqlite retention time is uint64 ( #10 )
2022-06-20 12:46:29 +02:00
3858d8a85b
feat: config for sqlite-only store ( #8 )
2022-06-20 11:55:39 +02:00
Jakub Sokołowski
dd76f2c212
status: lowering store capacity to 100k
...
https://github.com/status-im/infra-status/issues/7
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-05-26 21:11:51 +02:00
Jakub Sokołowski
fc001b2985
status: raise thresholds for alerts and recovery
...
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-05-24 23:12:26 +02:00
Jakub Sokołowski
867c63ecfa
status: lowering store capacity further to 200k
...
This cannot work due to container memory limits set to 70%:
https://github.com/status-im/infra-role-nim-waku/blob/c59aaf4c/defaults/main.yml#L53-L57
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-05-23 20:07:08 +02:00
Jakub Sokołowski
ccc074f24c
status: lower store capacity to half million
...
1 million is about 1.5 GB, which is a bit match for storing
entirely in memory, so I'm lowering to half of that for now.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-05-23 00:33:47 +02:00
Jakub Sokołowski
f29af66ecb
status: increase store capacity to 1 million
...
10k is kinda low with message rate of 150 unique envelopes per minute.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-05-19 20:08:17 +02:00
Jakub Sokołowski
943fb68c70
enable Discovery V5 for all status nodes
...
Issue: https://github.com/status-im/infra-status/issues/6
Depends: https://github.com/status-im/infra-role-nim-waku/commit/1e5367c8
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-03-18 12:46:18 +01:00
Jakub Sokołowski
37c3508180
status: specify node keys from BitWarden
...
https://github.com/status-im/infra-status/issues/1
In order to make sure we can re-create the hosts with the same address.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-03-11 18:32:35 +01:00
Jakub Sokołowski
b2142fd80b
add infra-role-waku-peers to connect peers
...
The logic was extracted to a separate role and Python script:
- https://github.com/status-im/infra-role-waku-peers
- https://github.com/status-im/infra-repos/commit/f3084d4e
- https://github.com/status-im/infra-role-nim-waku/commit/b8b64c0e
- https://github.com/status-im/infra-status/issues/4
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-03-10 17:26:44 +01:00
Jakub Sokołowski
4eac21225c
rename groups to be more generic, combine group_vars
...
This will make managing multiple fleets simpler.
Both prod and test files for nodes were identical.
Signed-off-by: Jakub Sokołowski <jakub@status.im>
2022-03-10 14:34:11 +01:00
Jakub Sokołowski
Hanno Cornelius