ceph crash archive-all
' ]% S1 ?4 \+ F1 h! `[root@node03 ~]# ceph health detail4 [2 O: b1 U Y" }, a
HEALTH_WARN 1 daemons have recently crashed
4 q# E. J& q; J' Q5 XRECENT_CRASH 1 daemons have recently crashed
% t$ P6 ^& }: r1 d; T$ ` client.rgw.node03 crashed on host node03 at 2023-01-06 07:12:28.313044Z7 \$ r- r# {" r7 A+ ?: R' q6 [
* r2 k5 V% Y8 e8 |* g# x+ \系统中所有的崩溃可以通过以下方式列出： a# I- x+ a' A/ j
[root@node03 ~]# ceph crash ls
4 P( B W* w) v5 u: h9 k4 n* eID ENTITY NEW , I& c+ G9 O0 t
2023-01-06_07:12:28.313044Z_df9f4fa8-d138-40f6-b8ec-c7c86299bb3e client.rgw.node03 * 8 f: h4 B4 D- p+ M- @
[root@node03 ~]# ceph crash ls-new8 @% c; X3 Z# W3 r
ID ENTITY NEW # e4 _; e- p/ f& U& }8 f! s" r6 f8 r
2023-01-06_07:12:28.313044Z_df9f4fa8-d138-40f6-b8ec-c7c86299bb3e client.rgw.node03 *
2 q' a! l2 R1 J+ U6 b& j# |5 z+ ?
) B+ P+ p5 U7 ]: U有关特定崩溃的信息可以通过以下方式检查' W# ?; ~% U/ ]4 S
[root@node03 ~]# ceph crash info 2023-01-06_07:12:28.313044Z_df9f4fa8-d138-40f6-b8ec-c7c86299bb3e! e2 t% u5 K% X+ `1 a! w
{
4 x: U) W$ c* ]2 d9 Y5 c) L0 X "os_version_id": "7.9",
! e0 P- ?4 x8 `- B' [3 g3 O6 E "utsname_release": "3.10.0-1160.el7.x86_64", ) {5 z) s2 W- C7 L" f/ h( ~% i S3 \% i
"os_name": "Red Hat Enterprise Linux Server",
; v' z& G) C3 C5 I" e/ t7 q "entity_name": "client.rgw.node03", D, ` a7 R; v: y0 M% I5 C
"timestamp": "2023-01-06 07:12:28.313044Z",
# c! ? q" n ^6 ^% e( i: A "process_name": "radosgw", + B& @, [& ~' U4 y" V
"utsname_machine": "x86_64",
6 t* R4 _# k7 w, s4 j$ X: B "utsname_sysname": "Linux",
6 C/ I6 `4 N* @2 B7 L1 c: u "os_version": "7.9 (Maipo)",
: a0 [; f j4 W+ H1 s1 b6 B* F* K$ ?9 Y "os_id": "rhel",
; s. ~- m% N: E+ ^ "utsname_version": "#1 SMP Tue Aug 18 14:50:17 EDT 2020",
' f2 D$ N& w7 w3 U9 T "backtrace": [1 g$ n6 y( c' {" S1 o, Z9 R2 C" V* N) s
"(()+0xf630) [0x7f5588ef4630]",
M# _/ Y6 A$ k) q$ a Y- Z# p "(()+0x84556) [0x7f5594e49556]", 1 N' u- G, e9 T: i# p/ D7 }% u, o
"(()+0x86b0b) [0x7f5594e4bb0b]", ( K8 j. Q3 h/ K& f/ o
"(RGWSI_Notify::unwatch(RGWSI_RADOS::Obj&, unsigned long)+0x2a) [0x5607b503643a]", & R1 V0 I2 A( D; |% v5 b
"(RGWWatcher::C_ReinitWatch::finish(int)+0x34) [0x5607b503bb44]", . l9 v, N4 f: ~1 C5 }: a( T$ h
"(Context::complete(int)+0x9) [0x5607b4b59489]", - z" C. }& x2 a; ^3 i! H6 D7 D* D
"(Finisher::finisher_thread_entry()+0x16f) [0x7f558c0b8fff]",
7 H9 `4 n- T7 m+ V "(()+0x7ea5) [0x7f5588eecea5]", 7 \" l8 ^0 _- O% G- H+ R
"(clone()+0x6d) [0x7f55883f596d]"
, c! m' J7 }3 p9 s: K$ V- l ], ) |4 {& Q2 M0 z8 `
"utsname_hostname": "node03",
' R7 s4 R; }' e "crash_id": "2023-01-06_07:12:28.313044Z_df9f4fa8-d138-40f6-b8ec-c7c86299bb3e", . U) f, V/ P! N# Q$ V' J( u7 H
"ceph_version": "14.2.8-111.el7"
5 a' g! g( ]9 o9 J% y& K" C}' T$ ]0 b5 A" A% {7 h5 x8 X
' z* F: y; `0 B7 X9 Q
可以通过“存档”崩溃（可能是在管理员检查之后）来消除此警告，从而不会生成此警告
; p) p) e* }' F" B2 b; M[root@node03 ~]# ceph crash archive 2023-01-06_07:12:28.313044Z_df9f4fa8-d138-40f6-b8ec-c7c86299bb3e & J9 c+ s [3 T4 w
' Y4 R2 p; j! u6 r, S) C4 G* Z4 [
7 o& b# ]9 a- O
4 {6 x: X9 J9 m- A2 i: V: x# G0 K通过ceph crash ls仍然可以看到已存档的崩溃，但不是ceph crash ls-new即可看到。 “recent”所指的时间段由选项mgr/crash/warn_recent_interval控制（默认值：两周）。 可以通过以下方式完全禁用这些警告： ceph config set mgr mgr/crash/warn_recent_interval 0 : t$ S$ |- e3 v, {: T4 h5 |
|
发表于 2023-1-9 15:00:11