I was check the code and i see these lines for that AIO timeout. case qpid::linearstore::journal::RHM_IORES_PAGE_AIOWAIT: if (++aio_sleep_cnt > MAX_AIO_SLEEPS) THROW_STORE_EXCEPTION("Timeout waiting for AIO in MessageStoreImpl::recoverMessages()"); ::usleep(AIO_SLEEP_TIME_US); break;
And these are the defaults #define MAX_AIO_SLEEPS 100000 // tot: ~1 sec #define AIO_SLEEP_TIME_US 10 // 0.01 ms RHM_IORES_PAGE_AIOWAIT, ///< IO operation suspended - next page is waiting for AIO. So does page got blocked and its waiting for page availability? Ram On Mon, Nov 5, 2018 at 8:00 PM rammohan ganapavarapu < rammohanga...@gmail.com> wrote: > > Actually we have upgraded from qpid-cpp 0.28 to 1.35 and after that we see > this message > > 2018-10-27 18:58:25 [Store] warning Linear Store: Journal > "<journal-name>": Bad record alignment found at fid=0x4605b offs=0x107680 > (likely journal overwrite boundary); 19 filler record(s) required. > 2018-10-27 18:58:25 [Store] notice Linear Store: Journal "<journal-name>": > Recover phase write: Wrote filler record: fid=0x4605b offs=0x107680 > 2018-10-27 18:58:25 [Store] notice Linear Store: Journal "<journal-name>": > Recover phase write: Wr... few more Recover phase logs > > It worked fine for a day and started throwing this message: > > 2018-10-28 12:27:01 [Store] critical Linear Store: Journal "<name>": > get_events() returned JERR_JCNTL_AIOCMPLWAIT; wmgr_status: wmgr: pi=25 pc=8 > po=0 aer=1 edac=TFFF ps=[-------------------------A------] > 2018-10-28 12:27:01 [Broker] warning Exchange <name> cannot deliver to > queue <queue-name>: Queue <queue-name>: MessageStoreImpl::store() failed: > jexception 0x0202 jcntl::handle_aio_wait() threw JERR_JCNTL_AIOCMPLWAIT: > Timeout waiting for AIOs to complete. > (/home/rganapavarapu/rpmbuild/BUILD/qpid-cpp-1.35.0/src/qpid/linearstore/MessageStoreImpl.cpp:1211) > 2018-10-28 12:27:01 [Broker] error Connection exception: framing-error: > Queue <queue-name>: MessageStoreImpl::store() failed: jexception 0x0202 > jcntl::handle_aio_wait() threw JERR_JCNTL_AIOCMPLWAIT: Timeout waiting for > AIOs to complete. > (/home/rganapavarapu/rpmbuild/BUILD/qpid-cpp-1.35.0/src/qpid/linearstore/MessageStoreImpl.cpp:1211) > 2018-10-28 12:27:01 [Protocol] error Connection > qpid.server-ip:5672-client-ip:44457 closed by error: Queue <queue-name>: > MessageStoreImpl::store() failed: jexception 0x0202 > jcntl::handle_aio_wait() threw JERR_JCNTL_AIOCMPLWAIT: Timeout waiting for > AIOs to complete. > (/home/rganapavarapu/rpmbuild/BUILD/qpid-cpp-1.35.0/src/qpid/linearstore/MessageStoreImpl.cpp:1211)(501) > 2018-10-28 12:27:01 [Protocol] error Connection > qpid.server-ip:5672-client-ip:44457 closed by error: illegal-argument: > Value for replyText is too large(320) > > Thanks, > Ram > > > On Mon, Nov 5, 2018 at 3:34 PM rammohan ganapavarapu < > rammohanga...@gmail.com> wrote: > >> No, local disk. >> >> On Mon, Nov 5, 2018 at 3:26 PM Gordon Sim <g...@redhat.com> wrote: >> >>> On 05/11/18 22:58, rammohan ganapavarapu wrote: >>> > Gordon, >>> > >>> > We are using java client 0.28 version and qpidd-cpp 1.35 version >>> > (qpid-cpp-server-1.35.0-1.el7.x86_64), i dont know at what scenario its >>> > happening but after i restart broker and if we wait for few days its >>> > happening again. From the above logs do you have any pointers to check? >>> >>> Are you using NFS? >>> >>> >>> >>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: users-unsubscr...@qpid.apache.org >>> For additional commands, e-mail: users-h...@qpid.apache.org >>> >>>