I have a CentOS8 SERVER which is running 1 years SQL Server 2019.
Today mssql-server suddenly crashed and dmesg show me this error.
I have update my sqlserver to mssql-server-15.0.4198.2-10.x86_64.rpm but it do not work.
I can use nohup /opt/mssql/bin/sqlservr & way to temporary start my sqlserver.
But who can Tell me how to fixed? This segfault.
This is Segfault messages:
[4802271.382426] (sqlservr)[21167]: segfault at 1f51 ip 00007fe69c8ca1ae sp 00007ffc68c1fb58 error 4 in ld-2.28.so[7fe69c8ae000+28000]
[4802271.558185] (sqlservr)[21202]: segfault at 1f51 ip 00007fe69c8ca1ae sp 00007ffc68c1fb58 error 4 in ld-2.28.so[7fe69c8ae000+28000]
[4802271.809011] (sqlservr)[21207]: segfault at 1f51 ip 00007fe69c8ca1ae sp 00007ffc68c1fb58 error 4 in ld-2.28.so[7fe69c8ae000+28000]
[4802272.059075] (sqlservr)[21211]: segfault at 1f51 ip 00007fe69c8ca1ae sp 00007ffc68c1fb58 error 4 in ld-2.28.so[7fe69c8ae000+28000]
[4802289.297274] (sqlservr)[30846]: segfault at 1f51 ip 00007fe69c8ca1ae sp 00007ffc68c1fb58 error 4 in ld-2.28.so[7fe69c8ae000+28000]
Related
Operation steps & amp; problem phenomenon
1, sysbench prepare, 100 tables, each table 100 million, 50 concurrency.
Try attempts to lower the concurrent number to 25 or the data volume to 5kw. Eventually, there are various memory reporting errors in the create secondary index link.
parameter:
Physical Server memory: 128GB
gs_guc reload -N all -I all -c "shared_buffers='30GB'"
gs_guc reload -N all -I all -c "max_process_memory='90GB'"
gs_guc reload -N all -I all -c "maintenance_work_mem='10GB'"
Report wrong phenomenon 1:
249FATAL: `sysbench.cmdline.call_command' function failed: ./oltp_common.lua:245: db_bulk_insert_next() failed
FATAL: PQexec() failed: 7 memory is temporarily unavailable
FATAL: failed query was: CREATE INDEX k_56 ON sbtest56(k)
FATAL: `sysbench.cmdline.call_command' function failed: ./oltp_common.lua:253: SQL error, errno = 0, state = 'YY006': memory is temporarily unavailable
Creating table 'sbtest76'...
Inserting 100000000 records into
Report wrong phenomenon 2:
Message from syslogd#testserver at Feb 23 10:19:45 ...
systemd:Caught , cannot fork for core dump: Cannot allocate memory
Report wrong phenomenon 3：
opengauss hitch。
Creating a secondary index on 'sbtest9'...
Segmentation fault (core dumped)
log 3：
could not fork new process for connection: Cannot allocate memory
could not fork new process for connection: Cannot allocate memory
I'm trying to build webrtc android app using org.webrtc:google-webrtc:1.0.32006 but sometimes I'm getting signal 11 (SIGSEGV), code 1 (SEGV_MAPERR) errors.
Description of errors:
backtrace:
#00 pc 00000000004d6f10 /data/app/com.xxx/split_config.arm64_v8a.apk!lib/arm64-v8a/libjingle_peerconnection_so.so (offset 0x6d5000) (Java_org_webrtc_PeerConnection_nativeClose+32)
#00 pc 0000000000648881 /data/app/com.xxx/oat/arm64/base.vdex
backtrace:
#00 pc 00000000002ebfa6 /data/app/com.xxx/split_config.armeabi_v7a.apk!lib/armeabi-v7a/libjingle_peerconnection_so.so (offset 0x435000)
#00 pc 0000000000038863 /data/app/com.xxx/oat/arm/base.odex (offset 0x38000)
The first error's keyword is Java_org_webrtc_PeerConnection_nativeClose but I couldn't find anything. Any idea of how to manage these errors and fix them?
I am installing k8s and vsphere CPI/CSI following the instructions located here
My setup:
2x centos 7.7 vSphere VM's (50g hd/16g ram), 1 master & 1 node in k8s cluster.
Made it to the part where I create the storageClass (near the end) when I discovered this github issue exactly. OP of the linked issue just started from scratch and their issue went away, so the report was closed. This has not been the case for me as I've redeployed my k8s cluster from scratch a bunch of times now and always hit this wall. Below is the error if you don't want to check the linked github issue.
Anyone have ideas on what I can try to get past this? I've checked my hd and ram and plenty there.
# kubectl -n kube-system logs pod/vsphere-csi-controller-0 vsphere-csi-controller
I0127 18:49:43.292667 1 config.go:261] GetCnsconfig called with cfgPath: /etc/cloud/csi-vsphere.conf
I0127 18:49:43.292859 1 config.go:206] Initializing vc server 132.250.31.180
I0127 18:49:43.292867 1 controller.go:67] Initializing CNS controller
I0127 18:49:43.292884 1 virtualcentermanager.go:63] Initializing defaultVirtualCenterManager...
I0127 18:49:43.292892 1 virtualcentermanager.go:65] Successfully initialized defaultVirtualCenterManager
I0127 18:49:43.292905 1 virtualcentermanager.go:107] Successfully registered VC "132.250.31.180"
I0127 18:49:43.292913 1 manager.go:60] Initializing volume.volumeManager...
I0127 18:49:43.292917 1 manager.go:64] volume.volumeManager initialized
time="2020-01-27T18:50:03Z" level=info msg="received signal; shutting down" signal=terminated
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x128 pc=0x867dc7]
goroutine 10 [running]:
google.golang.org/grpc.(*Server).GracefulStop(0x0)
/go/pkg/mod/google.golang.org/grpc#v1.23.0/server.go:1393 +0x37
github.com/rexray/gocsi.(*StoragePlugin).GracefulStop.func1()
/go/pkg/mod/github.com/rexray/gocsi#v1.0.0/gocsi.go:333 +0x35
sync.(*Once).Do(0xc0002cc8fc, 0xc000380ef8)
/usr/local/go/src/sync/once.go:44 +0xb3
github.com/rexray/gocsi.(*StoragePlugin).GracefulStop(0xc0002cc870, 0x21183a0, 0xc000056018)
/go/pkg/mod/github.com/rexray/gocsi#v1.0.0/gocsi.go:332 +0x56
github.com/rexray/gocsi.Run.func3()
/go/pkg/mod/github.com/rexray/gocsi#v1.0.0/gocsi.go:121 +0x4e
github.com/rexray/gocsi.trapSignals.func1(0xc00052a240, 0xc000426990, 0xc000426900)
/go/pkg/mod/github.com/rexray/gocsi#v1.0.0/gocsi.go:502 +0x143
created by github.com/rexray/gocsi.trapSignals
/go/pkg/mod/github.com/rexray/gocsi#v1.0.0/gocsi.go:487 +0x107
Ok turns out this SIGSEGV was a bug or something and it was caused by a network timeout, making this error kind of a red herring.
Details: My vsphere-csi-controller-0 pod was (and still is actually) unable to reach the vsphere server which caused the container in the pod to timeout and trigger this SIGSEV fault. The CSI contributers updated some libraries and the fault is now gone but the timeout remains. Timeout appears to be my problem and not related to CSI but that's a new question :)
If you want the details of what was fixed in the CSI check the github link in the question.
I am trying to run this code on a server (RedPitaya) as well as a client (ubuntu virtual machine).
The program returns the following error messages when it is run on the client with root privileges:
root#VirtualBox:/.../rp_remote_acquire# ./rp_remote_acquire -m 1 -a 192.169.1.100 -p 5000 -k 0 -c 0 -d 64
mmap scope io failed (non-fatal), 1
mmap scope ddr a failed (non-fatal), 1
Segmentation fault (core dumped)
I am not sure if the segmentation fault is related to the first two errors because I only get a segmentation fault when the server is running...
The error seems to be coming from here:
if (param->mapped_io == MAP_FAILED) {
fprintf(stderr, "mmap scope io failed (non-fatal), %d\n", errno);
param->mapped_io = NULL;
}
I am aware that a similar problem has already been resolved on stackoverflow.
I tried
sysctl dev.mem.restricted
and I tried adding
linux /boot/vmlinuz-linux iomem=relaxed
to the end of
/boot/grub/grub.cfg
and rebooting, but the problem still persists...
I would like to allow this program to access the computers virtual memory and thereby hopefully resolve all the errors. It could well be that I didn't manage to set the kernel parameters correctly.
Could someone please point me in the right direction?
I have installed OpenMPI and tried to compile/execute one of the examples delivered with the newest version.
As I try to run with mpiexec it says that the address is already in use.
Someone got a hint why this is always happening?
Kristians-MacBook-Pro:examples kristian$ mpicc -o hello hello_c.c
Kristians-MacBook-Pro:examples kristian$ mpiexec -n 4 ./hello
[Kristians-MacBook-Pro.local:02747] [[56076,0],0] bind() failed on error Address already in use (48)
[Kristians-MacBook-Pro.local:02747] [[56076,0],0] ORTE_ERROR_LOG: Error in file oob_usock_component.c at line 228
[Kristians-MacBook-Pro.local:02748] [[56076,1],0] usock_peer_send_blocking: send() to socket 19 failed: Socket is not connected (57)
[Kristians-MacBook-Pro.local:02748] [[56076,1],0] ORTE_ERROR_LOG: Unreachable in file oob_usock_connection.c at line 315
[Kristians-MacBook-Pro.local:02748] [[56076,1],0] orte_usock_peer_try_connect: usock_peer_send_connect_ack to proc [[56076,0],0] failed: Unreachable (-12)
[Kristians-MacBook-Pro.local:02749] [[56076,1],1] usock_peer_send_blocking: send() to socket 20 failed: Socket is not connected (57)
[Kristians-MacBook-Pro.local:02749] [[56076,1],1] ORTE_ERROR_LOG: Unreachable in file oob_usock_connection.c at line 315
[Kristians-MacBook-Pro.local:02749] [[56076,1],1] orte_usock_peer_try_connect: usock_peer_send_connect_ack to proc [[56076,0],0] failed: Unreachable (-12)
-------------------------------------------------------
Primary job terminated normally, but 1 process returned
a non-zero exit code.. Per user-direction, the job has been aborted.
-------------------------------------------------------
--------------------------------------------------------------------------
mpiexec detected that one or more processes exited with non-zero status, thus causing
the job to be terminated. The first process to do so was:
Process name: [[56076,1],0]
Exit code: 1
--------------------------------------------------------------------------
Thanks in advance.
Okay.
I have now changed the $TMPDIR environment variable with export TMPDIR=/tmp and it works.
Now it seems to me that the OpenMPI Session folder was blocking my communication. But why did it?
Am I missing something here?