Project

General

Profile

Actions

Bug #57990

closed

Crimson OSD crashes when trying to bring it up

Added by Luciano Lo Giudice over 1 year ago. Updated 17 days ago.

Status:
Closed
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

Hello,

Using the `crimson-osd` Ubuntu package for Quincy, we're seeing somewhat recurrent crashes when trying to bring up the crimson OSD by e.g using `crimson-osd -i $ID`.

The following backtrace was produced:

0# crimson::os::seastore::lba_manager::btree::btree_pin_set_t::add_pin(crimson::os::seastore::lba_manager::btree::btree_range_pin_t&) at /home/ubuntu/src/ceph/build/../src/crimson/os/seastore/lba_manager/btree/btree_range_pin.h:86
1# operator() at /home/ubuntu/src/ceph/build/../src/crimson/os/seastore/lba_manager/btree/lba_btree.cc:467
2# _ZN7seastar12continuationINS_8internal22promise_base_with_typeIN5boost13intrusive_ptrIN7crimson2os8seastore11lba_manager5btree11LBALeafNodeEEEEEZNS5_9erroratorIJNS5_19unthrowable_wrapperIRKSt10error_codeL_ZNS5_2ecILSt4errc5EEEEEEEE7_futureINS5_23errorated_future_markerISB_EEE9safe_thenIZNS5_13interruptible27interruptible_future_detailINS7_28TransactionConflictConditionESP_E23safe_then_interruptibleILb1EZNS9_8LBABtree13get_leaf_nodeENS9_12op_context_tENS7_7paddr_tEEUlSB_E_SB_Li0EEEDaOT0_EUlOSB_E_NSL_12pass_furtherEEEDaOT_S11_EUlS16_E_ZNS_6futureISB_E17then_wrapped_nrvoISU_S17_EENS_8futurizeIS15_E4typeES11_EUlOSC_RS17_ONS_12future_stateISB_EEE_SB_E15run_and_disposeEv at /home/ubuntu/src/ceph/build/../src/seastar/include/seastar/core/future.hh:606
3# seastar::reactor::run_tasks(seastar::reactor::task_queue&) at /home/ubuntu/src/ceph/build/../src/seastar/src/core/reactor.cc:2231
4# seastar::reactor::run_some_tasks() [clone .part.0] at /home/ubuntu/src/ceph/build/../src/seastar/src/core/reactor.cc:2640
5# seastar::reactor::do_run() at /home/ubuntu/src/ceph/build/../src/seastar/src/core/reactor.cc:2809
6# seastar::reactor::run() at /home/ubuntu/src/ceph/build/../src/seastar/src/core/reactor.cc:2697
7# seastar::app_template::run_deprecated(int, char**, std::function<void ()>&&) at /home/ubuntu/src/ceph/build/../src/seastar/src/core/app-template.cc:218
8# seastar::app_template::run(int, char**, std::function<seastar::future<int> ()>&&) at /home/ubuntu/src/ceph/build/../src/seastar/src/core/app-template.cc:125
9# main at /home/ubuntu/src/ceph/build/../src/crimson/osd/main.cc:239
10# __libc_start_main in /lib64/libc.so.6

It appears to be random, or at the very least very hard to reproduce. It's happened with different types of OSD, although every crash that we know of happened when using the seastar backend.

Actions

Also available in: Atom PDF