Page MenuHomePhabricator
Create Task
Maniphest T169289

Tool Labs 2017-06-29 Labstore100[45] kernel upgrade issues
Closed, ResolvedPublic
Actions

Description

Incident report on wikitech

Roughly what occurred (needs times and check for accuracy/order):

  • Rebooted labstore1004 activate upgraded kernel
  • Promoted labstore1004 to primary and failed clients over to it
  • Load spiked on labstore1004
  • (LDAP outage from unrelated causes)
  • Rebooted labstore1005 to activate upgraded kernel
  • Promoted labstore1005 to primary and failed clients over to it
  • Load spiked on labstore1005
  • Load spiked on several NFS clients
  • Reboots of various NFS clients in Tools project to ensure that stale NFS handles are not to blame for load spikes
  • Kubernetes nodes not able to communicate with etcd
  • Reboot flannel and flannel etcd
  • Reboot kubernetes etcd
  • Load continues to be very very high on labstore1005 NFS primary
  • Halt new pod scheduling on Kubernetes
  • Rolling reboot of Kubernetes nodes
  • Rolling reboot of grid engine nodes
  • Tune kernel parameters on labstore1005
  • Let things sit to see if load will settle down
  • Rollback labstore1004 kernel to 4.4.2-3+wmf8
  • Change i/o scheduler on labstore1005 from deadline to cfq
  • Let things sit to see if load will settle down
  • Re-enable new pod scheduling on Kubernetes to restore service to clients
  • Let things sit to see if load will settle down
  • 2017-06-30T00:28:26 Load spikes hit new high of 165.14 1m avg on labstore1005
  • Promote labstore1004 to NFS primary and fail clients over
  • Load on labstore1004 stays within pre-update expected values
  • Rollback labstore1005 kernel to 4.4.2-3+wmf8
  • Let things sit

Related Objects
Search...

StatusSubtypeAssignedTask
Restricted Task
Restricted Task
 Resolved• BstormT169289 Tool Labs 2017-06-29 Labstore100[45] kernel upgrade issues
 Resolved• BstormT169287 etcd config depends on puppet certs, but puppet doesn't know
 Resolved• BstormT169286 labstore1005 A PCIe link training failure error on boot
 ResolvedMoritzMuehlenhoffT169290 New anti-stackclash (4.9.25-1~bpo8+3 ) kernel super bad for NFS
 Resolved• BstormT203254 labstore1004 and labstore1005 high load issues following upgrades
 Resolved• BstormT224582 Migrate labstore1004/labstore1005 to Stretch/Buster
 Resolved• BstormT253353 Add cluster-awareness to nfs-exportd
 DeclinedNoneT257945 NFS v4.1/2 as possible fix for elevated load and lock contention on our NFS servers
 ResolvedaborreroT277653 Toolforge: add Debian Buster to the grid and eliminate Debian Stretch
 ResolvedaborreroT277866 cloud-init: figure out how to change /etc/hosts from cloud-init/vendordata
 ResolvedaborreroT278232 Toolforge: figure out how to work with the new domain in the grid
 ResolvedaborreroT278748 Toolforge: introduce support for selecting grid queue release
 ResolvedaborreroT282972 "--release is not implemented for --backend=kubernetes" with latest tools-webservice from Git
 ResolvedtaaviT288961 /usr/local/bin/webservice:109: DeprecationWarning: dist() and linux_distribution() functions are deprecated in Python 3.5
 ResolvedaborreroT300501 Toolforge grid: uwsgi in buster fails to load python3 venvs
 ResolvedtaaviT280037 Toolforge: set up monitoring tooling for stretch deprecation
 DuplicateNoneT280252 Toolforge Buster bastion no longer tab completes become command
 ResolvedtaaviT284767 Toolforge: migrate cron servers to Debian Buster
 DeclinedNoneT298089 Sort out Mono repositories for the buster grid
 ResolvedaborreroT298948 Toolforge grid deployment/management automation
 DeclinedNoneT300032 spicerack: introduce GridEngine controller
 ResolvedaborreroT301665 Toolforge jobs framework: create documentation on wikitech
 ResolvedtaaviT309525 Toolforge: Create a cookbook to decomission a SGE node
 ResolvedtaaviT309732 New SGE nodes can't talk to the grid engine master
 Resolved• nskaggsT309821 Buster webservice grid went BOOM!
 DeclinedtaaviT309902 Tiny swap on many grid nodes
 DeclinedNoneT336034 Toolforge grid automation: consider creating a cookbook to heal the grid from D state procs
 Resolved• BstormT169281 Labstore nfsd processes report "sent only x when sending y bytes - shutting down socket"

Event Timeline

• bd808 created this task.Jun 30 2017, 2:33 AM
Restricted Application added a project: Cloud-Services. · View Herald TranscriptJun 30 2017, 2:33 AM
Restricted Application added a subscriber: Aklapper. · View Herald Transcript
• bd808 added a subtask: T169287: etcd config depends on puppet certs, but puppet doesn't know.Jun 30 2017, 2:33 AM
• bd808 added a subtask: T169286: labstore1005 A PCIe link training failure error on boot.Jun 30 2017, 2:35 AM
Andrew added a subtask: T169290: New anti-stackclash (4.9.25-1~bpo8+3 ) kernel super bad for NFS.Jun 30 2017, 2:38 AM
• bd808 added a comment.Jun 30 2017, 2:38 AM

Some data about the system load we saw:

P5652 labstore1005 load following 2017-06-29 kernel upgrade
1bd808$ while /bin/true; do w|head -1; sleep 60; done
2 21:48:22 up 6:26, 5 users, load average: 22.48, 21.74, 24.36
3 21:49:22 up 6:27, 5 users, load average: 20.15, 21.54, 24.15
4 21:50:22 up 6:28, 5 users, load average: 15.01, 19.54, 23.28
5 21:51:22 up 6:29, 5 users, load average: 23.41, 21.89, 23.90
6 21:52:22 up 6:30, 5 users, load average: 9.67, 18.17, 22.49
7 21:53:22 up 6:31, 5 users, load average: 5.26, 15.47, 21.30
8 21:54:22 up 6:32, 5 users, load average: 3.68, 13.05, 20.09
9 21:55:22 up 6:33, 5 users, load average: 9.31, 12.55, 19.46
10 21:56:22 up 6:34, 5 users, load average: 22.45, 16.02, 20.23
11 21:57:22 up 6:35, 5 users, load average: 16.71, 15.48, 19.78
12 21:58:22 up 6:36, 5 users, load average: 10.81, 14.23, 19.09
13 21:59:22 up 6:37, 5 users, load average: 4.34, 11.72, 17.92
14 22:00:22 up 6:38, 5 users, load average: 20.87, 13.74, 18.15
15 22:01:22 up 6:39, 5 users, load average: 107.12, 40.87, 27.27
16 22:02:22 up 6:40, 5 users, load average: 79.89, 46.28, 30.06
17 22:03:22 up 6:41, 5 users, load average: 39.44, 40.80, 29.19
18 22:04:22 up 6:42, 5 users, load average: 23.71, 35.94, 28.24
19 22:05:22 up 6:43, 5 users, load average: 23.53, 33.20, 27.75
20 22:06:22 up 6:44, 5 users, load average: 23.09, 31.63, 27.57
21 22:07:22 up 6:45, 5 users, load average: 13.49, 27.30, 26.33
22 22:08:22 up 6:46, 5 users, load average: 17.68, 25.54, 25.76
23 22:09:22 up 6:47, 5 users, load average: 18.33, 24.19, 25.28
24 22:10:23 up 6:48, 5 users, load average: 33.64, 26.41, 25.92
25 22:11:23 up 6:49, 5 users, load average: 42.27, 31.18, 27.65
26 22:12:23 up 6:50, 5 users, load average: 28.28, 29.27, 27.22
27 22:13:23 up 6:51, 5 users, load average: 23.89, 28.23, 27.01
28 22:14:23 up 6:52, 6 users, load average: 19.23, 26.27, 26.42
29 22:15:23 up 6:53, 6 users, load average: 29.90, 27.27, 26.71
30 # <chasemp> !log set cfq scheduler on labstore1005
31 22:16:23 up 6:54, 6 users, load average: 39.06, 30.90, 28.03
32 22:17:23 up 6:55, 6 users, load average: 29.66, 29.78, 27.83
33 22:18:23 up 6:56, 6 users, load average: 28.39, 29.13, 27.72
34 22:19:23 up 6:57, 6 users, load average: 19.96, 26.63, 26.95
35 22:20:23 up 6:58, 6 users, load average: 17.86, 24.48, 26.17
36 22:21:23 up 6:59, 6 users, load average: 31.95, 27.74, 27.21
37 22:22:23 up 7:00, 6 users, load average: 24.14, 26.34, 26.77
38 22:23:23 up 7:01, 6 users, load average: 18.88, 24.37, 26.05
39 22:24:23 up 7:02, 6 users, load average: 24.18, 24.08, 25.80
40 22:25:23 up 7:03, 6 users, load average: 40.34, 29.36, 27.55
41 22:26:23 up 7:04, 6 users, load average: 36.21, 30.58, 28.11
42 22:27:23 up 7:05, 6 users, load average: 22.02, 27.42, 27.17
43 22:28:23 up 7:06, 6 users, load average: 17.21, 24.87, 26.29
44 22:29:23 up 7:07, 6 users, load average: 9.84, 21.39, 25.01
45 22:30:23 up 7:08, 6 users, load average: 15.81, 20.20, 24.33
46 22:31:23 up 7:09, 6 users, load average: 42.33, 27.60, 26.66
47 22:32:23 up 7:10, 6 users, load average: 24.94, 25.29, 25.92
48 22:33:23 up 7:11, 6 users, load average: 15.22, 22.62, 24.98
49 22:34:23 up 7:12, 6 users, load average: 6.63, 18.73, 23.48
50 22:35:23 up 7:13, 6 users, load average: 11.32, 17.40, 22.70
51 22:36:23 up 7:14, 6 users, load average: 20.44, 19.36, 23.07
52 22:37:23 up 7:15, 6 users, load average: 9.59, 16.41, 21.82
53 22:38:23 up 7:16, 6 users, load average: 5.33, 13.94, 20.63
54 22:39:23 up 7:17, 6 users, load average: 4.71, 12.06, 19.56
55 22:40:23 up 7:18, 6 users, load average: 18.87, 14.60, 19.95
56 22:41:23 up 7:19, 6 users, load average: 18.17, 15.34, 19.88
57 22:42:23 up 7:20, 6 users, load average: 17.02, 15.53, 19.66
58 22:43:23 up 7:21, 6 users, load average: 9.19, 13.71, 18.78
59 22:44:23 up 7:22, 6 users, load average: 7.81, 12.37, 18.00
60 22:45:23 up 7:23, 6 users, load average: 15.55, 13.35, 17.96
61 22:46:23 up 7:24, 6 users, load average: 29.23, 17.72, 19.18
62 22:47:23 up 7:25, 6 users, load average: 22.96, 18.22, 19.27
63 22:48:24 up 7:26, 6 users, load average: 18.69, 17.75, 19.04
64 22:49:24 up 7:27, 6 users, load average: 17.10, 17.72, 18.96
65 22:50:24 up 7:28, 6 users, load average: 12.75, 16.20, 18.35
66 22:51:24 up 7:29, 6 users, load average: 18.27, 17.72, 18.77
67 22:52:24 up 7:30, 6 users, load average: 10.41, 15.61, 17.98
68 22:53:24 up 7:31, 6 users, load average: 6.43, 13.67, 17.18
69 22:54:24 up 7:32, 6 users, load average: 4.77, 11.82, 16.31
70 22:55:24 up 7:33, 6 users, load average: 12.29, 12.35, 16.20
71 22:56:24 up 7:34, 6 users, load average: 19.81, 14.69, 16.77
72 22:57:24 up 7:35, 6 users, load average: 15.23, 14.24, 16.48
73 22:58:24 up 7:36, 6 users, load average: 7.96, 12.50, 15.75
74 22:59:24 up 7:37, 6 users, load average: 4.49, 10.58, 14.88
75 # Re-enabled Kubernetes scheduler
76 23:00:24 up 7:38, 6 users, load average: 30.33, 15.23, 16.12
77 23:01:24 up 7:39, 6 users, load average: 77.63, 32.98, 22.25
78 23:02:24 up 7:40, 6 users, load average: 49.11, 33.68, 23.22
79 23:03:24 up 7:41, 6 users, load average: 31.96, 31.74, 23.21
80 23:04:24 up 7:42, 6 users, load average: 23.84, 29.24, 22.87
81 23:05:24 up 7:43, 6 users, load average: 25.82, 28.46, 22.99
82 23:06:24 up 7:44, 6 users, load average: 26.25, 28.69, 23.44
83 23:07:24 up 7:45, 6 users, load average: 14.66, 24.97, 22.49
84 23:08:24 up 7:46, 6 users, load average: 13.73, 22.73, 21.87
85 23:09:24 up 7:47, 6 users, load average: 9.77, 20.01, 20.99
86 23:10:24 up 7:48, 6 users, load average: 25.70, 21.49, 21.37
87 23:11:24 up 7:49, 6 users, load average: 24.64, 22.67, 21.83
88 23:12:24 up 7:50, 6 users, load average: 16.08, 20.23, 21.02
89 23:13:24 up 7:51, 6 users, load average: 15.49, 19.19, 20.61
90 23:14:24 up 7:52, 6 users, load average: 13.61, 18.20, 20.19
91 23:15:24 up 7:53, 6 users, load average: 22.57, 19.67, 20.55
92 23:16:24 up 7:54, 6 users, load average: 35.45, 23.99, 21.99
93 23:17:24 up 7:55, 6 users, load average: 29.25, 23.99, 22.10
94 23:18:24 up 7:56, 6 users, load average: 31.99, 25.49, 22.72
95 23:19:24 up 7:57, 6 users, load average: 39.21, 28.43, 23.89
96 23:20:24 up 7:58, 6 users, load average: 36.44, 29.37, 24.49
97 23:21:24 up 7:59, 6 users, load average: 43.42, 32.92, 26.03
98 23:22:24 up 8:00, 6 users, load average: 43.83, 35.04, 27.19
99 23:23:24 up 8:01, 6 users, load average: 30.10, 32.89, 26.96
100 23:24:24 up 8:02, 6 users, load average: 35.34, 33.66, 27.58
101 23:25:24 up 8:03, 6 users, load average: 53.45, 38.55, 29.60
102 23:26:25 up 8:04, 6 users, load average: 42.10, 38.54, 30.19
103 23:27:25 up 8:05, 6 users, load average: 27.22, 34.98, 29.49
104 23:28:25 up 8:06, 6 users, load average: 19.43, 31.20, 28.52
105 23:29:25 up 8:07, 6 users, load average: 16.17, 28.22, 27.67
106 23:30:25 up 8:08, 6 users, load average: 34.74, 30.24, 28.34
107 23:31:25 up 8:09, 6 users, load average: 55.58, 37.68, 31.05
108 23:32:25 up 8:10, 6 users, load average: 38.48, 36.08, 30.92
109 23:33:25 up 8:11, 6 users, load average: 30.98, 34.18, 30.58
110 23:34:25 up 8:12, 6 users, load average: 24.63, 31.89, 30.03
111 23:36:25 up 8:14, 6 users, load average: 34.26, 33.03, 30.64
112 23:37:25 up 8:15, 6 users, load average: 24.90, 30.52, 29.93
113 23:38:25 up 8:16, 6 users, load average: 21.08, 28.21, 29.16
114 23:39:25 up 8:17, 6 users, load average: 15.94, 25.44, 28.14
115 23:40:25 up 8:18, 6 users, load average: 23.06, 25.17, 27.85
116 23:41:25 up 8:19, 6 users, load average: 25.25, 26.07, 28.03
117 23:42:25 up 8:20, 6 users, load average: 20.00, 24.12, 27.22
118 23:43:25 up 8:21, 6 users, load average: 15.96, 22.19, 26.36
119 23:44:25 up 8:22, 6 users, load average: 23.61, 23.32, 26.49
120 23:45:25 up 8:23, 6 users, load average: 30.87, 24.97, 26.83
121 23:46:25 up 8:24, 6 users, load average: 37.74, 28.79, 28.07
122 23:47:25 up 8:25, 6 users, load average: 29.97, 27.89, 27.79
123 23:48:25 up 8:26, 6 users, load average: 29.88, 28.12, 27.86
124 23:49:25 up 8:27, 6 users, load average: 31.49, 28.37, 27.93
125 23:50:25 up 8:28, 6 users, load average: 24.30, 27.03, 27.51
126 23:51:25 up 8:29, 6 users, load average: 20.32, 25.68, 27.03
127 23:52:25 up 8:30, 6 users, load average: 17.82, 23.77, 26.28
128 23:53:25 up 8:31, 6 users, load average: 15.61, 22.18, 25.58
129 23:54:25 up 8:32, 6 users, load average: 14.75, 20.46, 24.76
130 23:55:25 up 8:33, 6 users, load average: 30.94, 23.56, 25.52
131 23:56:25 up 8:34, 6 users, load average: 37.85, 27.34, 26.73
132 23:57:25 up 8:35, 6 users, load average: 31.53, 27.32, 26.75
133 23:58:25 up 8:36, 6 users, load average: 21.11, 25.36, 26.13
134 23:59:25 up 8:37, 5 users, load average: 11.26, 21.81, 24.86
135 00:00:25 up 8:38, 5 users, load average: 30.35, 24.46, 25.53
136 00:01:25 up 8:39, 5 users, load average: 128.66, 53.90, 35.60
137 00:02:26 up 8:40, 5 users, load average: 118.68, 65.93, 40.96
138 00:03:26 up 8:41, 5 users, load average: 68.00, 61.41, 41.01
139 00:04:26 up 8:42, 5 users, load average: 48.45, 57.09, 40.80
140 00:05:26 up 8:43, 5 users, load average: 30.11, 50.17, 39.43
141 00:06:26 up 8:44, 5 users, load average: 37.95, 48.86, 39.66
142 00:07:26 up 8:45, 5 users, load average: 25.56, 43.11, 38.25
143 00:08:26 up 8:46, 5 users, load average: 17.92, 38.16, 36.88
144 00:09:26 up 8:47, 4 users, load average: 11.29, 32.67, 35.07
145 00:10:26 up 8:48, 4 users, load average: 32.73, 33.91, 35.29
146 00:11:26 up 8:49, 4 users, load average: 35.13, 34.68, 35.48
147 00:12:26 up 8:50, 4 users, load average: 31.50, 33.41, 34.98
148 00:13:26 up 8:51, 4 users, load average: 26.91, 31.79, 34.33
149 00:14:26 up 8:52, 4 users, load average: 29.68, 31.24, 33.95
150 00:15:26 up 8:53, 4 users, load average: 33.28, 32.27, 34.15
151 00:16:26 up 8:54, 4 users, load average: 42.20, 34.76, 34.88
152 00:17:26 up 8:55, 4 users, load average: 26.46, 31.80, 33.86
153 00:18:26 up 8:56, 4 users, load average: 24.92, 30.23, 33.18
154 00:19:26 up 8:57, 4 users, load average: 21.84, 28.31, 32.33
155 00:20:26 up 8:58, 4 users, load average: 27.89, 28.66, 32.19
156 00:21:26 up 8:59, 4 users, load average: 33.64, 30.41, 32.58
157 00:22:26 up 9:00, 4 users, load average: 30.08, 30.01, 32.31
158 00:23:26 up 9:01, 4 users, load average: 23.74, 28.01, 31.47
159 00:24:26 up 9:02, 4 users, load average: 51.83, 35.55, 33.86
160 00:25:26 up 9:03, 4 users, load average: 92.31, 48.55, 38.34
161 00:26:26 up 9:04, 4 users, load average: 124.97, 67.08, 45.42
162 00:27:26 up 9:05, 4 users, load average: 126.32, 76.00, 49.73
163 00:28:26 up 9:06, 4 users, load average: 165.14, 96.48, 58.43
164 00:29:26 up 9:07, 4 users, load average: 132.22, 100.43, 62.23
165 00:30:26 up 9:08, 4 users, load average: 98.44, 96.11, 63.12
166 00:31:26 up 9:09, 4 users, load average: 113.44, 99.98, 66.47
167 00:32:26 up 9:10, 4 users, load average: 91.29, 96.83, 67.53
168 00:33:26 up 9:11, 4 users, load average: 67.54, 88.47, 66.46
169 00:34:26 up 9:12, 4 users, load average: 90.20, 91.37, 68.85
170 00:35:26 up 9:13, 4 users, load average: 66.07, 84.52, 67.92
171 00:36:26 up 9:14, 4 users, load average: 65.21, 81.17, 67.82
172 00:37:27 up 9:15, 4 users, load average: 37.02, 70.30, 64.92
173 00:38:27 up 9:16, 4 users, load average: 23.21, 60.12, 61.75
174 00:39:27 up 9:17, 4 users, load average: 14.69, 50.88, 58.46
175 00:40:27 up 9:18, 4 users, load average: 18.34, 44.83, 55.88
176 00:41:27 up 9:19, 4 users, load average: 24.19, 42.21, 54.31
177 00:42:27 up 9:20, 4 users, load average: 19.45, 37.63, 51.98
178 00:43:27 up 9:21, 4 users, load average: 39.88, 40.26, 51.99
179 00:44:27 up 9:22, 4 users, load average: 35.88, 38.82, 50.74
180 00:45:27 up 9:23, 4 users, load average: 30.76, 36.33, 49.12
181 00:46:27 up 9:24, 4 users, load average: 50.64, 41.56, 50.14
182 00:47:27 up 9:25, 4 users, load average: 53.48, 44.15, 50.50
183 00:48:27 up 9:26, 4 users, load average: 50.56, 44.74, 50.29
184 00:49:27 up 9:27, 4 users, load average: 35.02, 41.29, 48.75
185 00:50:27 up 9:28, 4 users, load average: 38.36, 40.51, 47.98
186 00:51:27 up 9:29, 4 users, load average: 46.91, 42.62, 48.25
187 00:52:27 up 9:30, 4 users, load average: 28.36, 38.41, 46.47
188 00:53:27 up 9:31, 4 users, load average: 20.75, 34.56, 44.65
189 00:54:27 up 9:32, 4 users, load average: 29.30, 34.09, 43.83
190 00:55:27 up 9:33, 4 users, load average: 41.25, 36.48, 44.02
191 00:56:27 up 9:34, 4 users, load average: 36.59, 36.44, 43.56
192 00:57:27 up 9:35, 4 users, load average: 28.67, 34.31, 42.39
193 00:58:27 up 9:36, 4 users, load average: 36.35, 35.43, 42.26
194 00:59:27 up 9:37, 4 users, load average: 39.51, 36.75, 42.30
195 01:00:27 up 9:38, 4 users, load average: 53.77, 39.30, 42.72

• bd808 added a comment.Jun 30 2017, 2:56 AM

SAL entries:

== 2017-06-30 ==
02:29	<chasemp>	labstore1005 start drbd
02:14	<chasemp>	reboot labstore1005 (5m ago)
01:33	<chasemp>	time for i in `cat tools-hosts`; do ssh -i ~/.ssh/labs_root_id_rsa root@$i.eqiad.wmflabs 'hostname -f; uptime; tc-setup'; done
01:29	<andrewbogott>	rebooting tools-cron-01
01:25	<chasemp>	reboot labstoer1005
01:23	<chasemp>	fail nfs from labstore1005 to labstore1004 (I failed to log a previous failover to 1004 and back)

== 2017-06-29 ==
22:16	<chasemp>	set cfq scheduler on labstore1005
21:40	<chasemp>	reboot labstore1004 with grub set to gnulinux-advanced-1773f282-5a1b-441e-865c-8b70a0ebc925>gnulinux-4.4.0-3-amd64-advanced-1773f282-5a1b-441e-865c-8b70a0ebc925
20:33	<andrewbogott>	depooling, rebooting, and repooling every lighttpd node three at a time
18:30	<chasemp>	restart nfs on labstore1004 (primary)
15:38	<chasemp>	restart nfs-exportd on labstore1004
17:22	<bd808>	rebooting tools-static-11
17:20	<andrewbogott>	rebooting tools-static-10
16:27	<chasemp>	restart k8s components on master (madhu)
16:10	<chasemp>	tools-flannel-etcd-01:~$ sudo service etcd restart
15:57	<chasemp>	reboot tools-docker-registery-01 for nfs
15:09	<chasemp>	set downtimes for labstore1004/1005 failover see https://etherpad.wikimedia.org/p/labstore_reboots
• bd808 added a subtask: T169281: Labstore nfsd processes report "sent only x when sending y bytes - shutting down socket".Jun 30 2017, 2:58 AM
• bd808 updated the task description. (Show Details)Jun 30 2017, 3:13 AM
• bd808 updated the task description. (Show Details)Jun 30 2017, 3:17 AM
Jay8g subscribed.Jun 30 2017, 4:07 AM
• madhuvishy mentioned this in T168584: Labsdb* servers need to be rebooted.Jun 30 2017, 3:06 PM
herron subscribed.Jun 30 2017, 5:46 PM
• chasemp triaged this task as High priority.Jun 30 2017, 6:15 PM
• madhuvishy subscribed.Jul 1 2017, 6:47 AM

Incident documentation - https://wikitech.wikimedia.org/wiki/Incident_documentation/20170629-Labstore_Kernel_Upgrade

• bd808 updated the task description. (Show Details)Jul 1 2017, 6:37 PM
Paladox subscribed.Jul 1 2017, 6:45 PM
• bd808 mentioned this in T144153: Move kubernetes authentication to using X.509 client certs.Jul 1 2017, 8:58 PM
MoritzMuehlenhoff closed subtask T169290: New anti-stackclash (4.9.25-1~bpo8+3 ) kernel super bad for NFS as Resolved.Sep 10 2018, 1:20 PM
• bd808 lowered the priority of this task from High to Medium.Nov 21 2018, 11:14 PM

Pretty sure this is not "High" priority 18 months later.

• GTirloni edited projects, added cloud-services-team (Kanban); removed Cloud-Services.Mar 23 2019, 9:30 PM
• Bstorm closed subtask T169287: etcd config depends on puppet certs, but puppet doesn't know as Resolved.Aug 14 2019, 5:39 PM
• Bstorm moved this task from Inbox to Graveyard on the cloud-services-team (Kanban) board.Nov 27 2019, 4:58 PM
• Bstorm subscribed.

Moving this to the graveyard and linking it to a more recent task.

• Bstorm added a subtask: T224582: Migrate labstore1004/labstore1005 to Stretch/Buster.Nov 27 2019, 4:59 PM
• Bstorm changed the status of subtask T224582: Migrate labstore1004/labstore1005 to Stretch/Buster from Open to Stalled.Mar 16 2020, 7:31 PM
• JHedden changed the status of subtask T169286: labstore1005 A PCIe link training failure error on boot from Open to Stalled.May 5 2020, 4:30 PM
• Bstorm changed the status of subtask T224582: Migrate labstore1004/labstore1005 to Stretch/Buster from Stalled to Open.May 19 2020, 9:10 PM
• Bstorm added a comment.May 21 2020, 10:10 PM

labstore1005 is now on Debian stretch. When we fail over to it next week, it will be interesting to see how load looks.

• Bstorm closed subtask T224582: Migrate labstore1004/labstore1005 to Stretch/Buster as Resolved.Jun 11 2020, 7:35 PM
• Bstorm closed this task as Resolved.Jun 11 2020, 11:04 PM
• Bstorm claimed this task.

Both labstores are now stretch. Overall, what was seen on this ticket has not happened. Load is higher in general on the servers, but it has no discernible impact on actual performance. Beyond that the clients are not seeing load spikes.

Since the kernel version is far beyond the version in this ticket, I don't think it is useful to keep this open.

• Bstorm mentioned this in T203254: labstore1004 and labstore1005 high load issues following upgrades.Jun 11 2020, 11:59 PM
• Bstorm closed subtask T169281: Labstore nfsd processes report "sent only x when sending y bytes - shutting down socket" as Resolved.Jun 22 2020, 11:32 PM
• nskaggs closed subtask T169286: labstore1005 A PCIe link training failure error on boot as Resolved.Jan 5 2022, 7:26 PM
Content licensed under Creative Commons Attribution-ShareAlike (CC BY-SA) 4.0 unless otherwise noted; code licensed under GNU General Public License (GPL) 2.0 or later and other open source licenses. By using this site, you agree to the Terms of Use, Privacy Policy, and Code of Conduct. · Wikimedia Foundation · Privacy Policy · Code of Conduct · Terms of Use · Disclaimer · CC-BY-SA · GPL · Credits