|
| 1 | +- commits: |
| 2 | + - subject: Update balena-supervisor to v17.1.2 |
| 3 | + hash: 6744c3312b70dd0fa3b87ca4dcd3a4efd7405680 |
| 4 | + body: | |
| 5 | + Update balena-supervisor from 17.0.3 to 17.1.2 |
| 6 | + footer: |
| 7 | + Change-type: patch |
| 8 | + change-type: patch |
| 9 | + author: balena-renovate[bot] |
| 10 | + nested: |
| 11 | + - commits: |
| 12 | + - subject: "config/backends/power-fan: Enable configurable power and fan for |
| 13 | + Forecr DSBOARD ORNX Nano 8GB" |
| 14 | + hash: 69b451c650a6d641a77b293abfd253d17db8a2ce |
| 15 | + body: "" |
| 16 | + footer: |
| 17 | + Change-type: patch |
| 18 | + change-type: patch |
| 19 | + Signed-off-by: Alexandru Costache < [email protected]> |
| 20 | + signed-off-by: Alexandru Costache < [email protected]> |
| 21 | + author: Alexandru Costache |
| 22 | + nested: [] |
| 23 | + version: balena-supervisor-17.1.2 |
| 24 | + title: "" |
| 25 | + date: 2025-07-28T14:08:15.142Z |
| 26 | + - commits: |
| 27 | + - subject: Remove memory healthcheck |
| 28 | + hash: 8ffdba7d187a00d846d2c76c3416a45e02bf0e93 |
| 29 | + body: > |
| 30 | + Supervisor has had memory leaks removed since v16.5.1, with |
| 31 | + latest tested |
| 32 | + |
| 33 | + version being v16.7.1. Furthermore, on recent reported instances |
| 34 | + of memory healthcheck |
| 35 | + |
| 36 | + triggering on support, we've snapshotted the heap before & after |
| 37 | + on devices multiple times |
| 38 | + |
| 39 | + without finding any evidence of memory leaks in the snapshots. |
| 40 | + |
| 41 | + |
| 42 | + Therefore, it's hypothesized that the heuristic for determining |
| 43 | + starting memory may |
| 44 | + |
| 45 | + be flawed in that it's not waiting long enough after system |
| 46 | + startup, or it may |
| 47 | + |
| 48 | + run right after garbage collection has happened. Because of the |
| 49 | + variability and |
| 50 | + |
| 51 | + difficulty of ascertaining these factors, we suspect an |
| 52 | + inaccurate memory |
| 53 | + |
| 54 | + baseline may be the cause of the instances of false positives on |
| 55 | + support. |
| 56 | + footer: |
| 57 | + See: https://balena.zulipchat.com/#narrow/channel/403752-channel.2Fsupport-help/topic/supervisor.20memory.20usage.20above.20threadhold/near/520640885 |
| 58 | + see: https://balena.zulipchat.com/#narrow/channel/403752-channel.2Fsupport-help/topic/supervisor.20memory.20usage.20above.20threadhold/near/520640885 |
| 59 | + Change-type: patch |
| 60 | + change-type: patch |
| 61 | + Signed-off-by: Christina Ying Wang < [email protected]> |
| 62 | + signed-off-by: Christina Ying Wang < [email protected]> |
| 63 | + author: Christina Ying Wang |
| 64 | + nested: [] |
| 65 | + version: balena-supervisor-17.1.1 |
| 66 | + title: "" |
| 67 | + date: 2025-06-09T13:58:34.068Z |
| 68 | + - commits: |
| 69 | + - subject: Support target state apply cancellation |
| 70 | + hash: 8c691662712d1b93264260bf6c7ded8d51bd7d55 |
| 71 | + body: > |
| 72 | + The current target state apply is cancelled when either: |
| 73 | + |
| 74 | + - /v1/update is called with cancel: true |
| 75 | + |
| 76 | + - A different target state is received from the cloud (with a |
| 77 | + non-304 status) |
| 78 | + |
| 79 | + |
| 80 | + Following apply cancellation, a target state apply is |
| 81 | + re-triggered. This ensures |
| 82 | + |
| 83 | + that the user can force a device out of a dead-locked situation |
| 84 | + where a long-running |
| 85 | + |
| 86 | + task such as an image fetch fails to cede control back to the |
| 87 | + Supervisor, which is |
| 88 | + |
| 89 | + the behavior observed in an Engine bug with infinite pull |
| 90 | + retries with a bad network. |
| 91 | + footer: |
| 92 | + Change-type: minor |
| 93 | + change-type: minor |
| 94 | + Signed-off-by: Christina Ying Wang < [email protected]> |
| 95 | + signed-off-by: Christina Ying Wang < [email protected]> |
| 96 | + author: Christina Ying Wang |
| 97 | + nested: [] |
| 98 | + version: balena-supervisor-17.1.0 |
| 99 | + title: "" |
| 100 | + date: 2025-05-28T15:42:41.610Z |
| 101 | + - commits: |
| 102 | + - subject: Exclude reclaimable slab memory from used memory metric |
| 103 | + hash: 49e91a2639fbee6333fbf3ea52d8d6a7acfd705d |
| 104 | + body: > |
| 105 | + Aligns metric value with used memory reported by the free and |
| 106 | + htop |
| 107 | + |
| 108 | + utilities. |
| 109 | + footer: |
| 110 | + Change-type: patch |
| 111 | + change-type: patch |
| 112 | + author: Ken Bannister |
| 113 | + nested: [] |
| 114 | + version: balena-supervisor-17.0.5 |
| 115 | + title: "" |
| 116 | + date: 2025-05-27T15:32:21.540Z |
| 117 | + - commits: |
| 118 | + - subject: Remove unsupported fields from contract requirements |
| 119 | + hash: 4318272844c7067288675ca6a2b0f4d0175c3ae1 |
| 120 | + body: > |
| 121 | + A contract including extra requirement fields, such as "name" |
| 122 | + would fail |
| 123 | + |
| 124 | + validation. This PR removes any extra fields from the validated |
| 125 | + contract |
| 126 | + |
| 127 | + to prevent services with these extra fields from getting |
| 128 | + rejected by the |
| 129 | + |
| 130 | + contract validation. |
| 131 | + footer: |
| 132 | + Change-type: patch |
| 133 | + change-type: patch |
| 134 | + author: Felipe Lalanne |
| 135 | + nested: [] |
| 136 | + version: balena-supervisor-17.0.4 |
| 137 | + title: "" |
| 138 | + date: 2025-05-22T18:47:27.797Z |
| 139 | + version: 6.6.9 |
| 140 | + title: "" |
| 141 | + date: 2025-09-29T17:46:37.294Z |
1 | 142 | - commits: |
2 | 143 | - subject: Update tests/leviathan digest to 73d4851 |
3 | 144 | hash: d0c870cd02f1ba1a2ddaa50d223160665fc056a1 |
|
0 commit comments