Hi Mike!
1. I'm able to reproduce spinning on guacd 0.9.12-incubating on docker,
by connecting to vnc server and logging out without disconnecting. If I
do this couple of times a row - then spinning occur.
2. I tried do the same with guacd 0.9.12-incubating on another machine
outside docker without success, but the difference is that:
a) (1) (with docker) uses postgresql for authentication
b) (2) (outside docker) uses legacy user-mapping xml file
3. I noticed some strange "udp6" connection between guacamole and
postgres (see below)
4. I also runned some diagnostic commands which you may find useful:
lala:~# top
top - 11:53:18 up 104 days, 22:37, 2 users, load average: 1,45, 1,43, 1,37
Tasks: 473 total, 1 running, 472 sleeping, 0 stopped, 0 zombie
%Cpu(s): 25,4 us, 0,2 sy, 0,0 ni, 74,4 id, 0,0 wa, 0,0 hi, 0,0 si,
0,0 st
KiB Mem: 32714444 total, 27957256 used, 4757188 free, 1023880 buffers
KiB Swap: 1048572 total, 55568 used, 993004 free. 19337388 cached Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
5796 root 20 0 495456 45820 9360 S 100,0 0,1 81:14.92 guacd
-------------------------------------
lala:~# netstat -pn | grep gua
tcp 0 0 192.168.1.2:5432 172.17.0.3:50100 ESTABLISHED
5752/postgres: guac
tcp 0 0 192.168.1.2:5432 172.17.0.3:50070 ESTABLISHED
5703/postgres: guac
udp6 0 0 ::1:48362 ::1:48362 ESTABLISHED
5703/postgres: guac
---------------------------------
lala:~# cat /proc/5796/stack
[<ffffffff810d37ec>] futex_wait_queue_me+0xcc/0x130
[<ffffffff810d3ac2>] futex_wait+0x162/0x260
[<ffffffff810798d4>] get_signal_to_deliver+0x104/0x5d0
[<ffffffff810d5c52>] do_futex+0xf2/0xb60
[<ffffffff810797a5>] ptrace_notify+0x55/0x80
[<ffffffff81021d2f>] tracehook_report_syscall_exit+0x2f/0xa0
[<ffffffff810d672e>] SyS_futex+0x6e/0x150
[<ffffffff81013f08>] do_notify_resume+0x78/0xa0
[<ffffffff8151ae4d>] system_call_fast_compare_end+0x10/0x15
[<ffffffffffffffff>] 0xffffffffffffffff
----------------------------------
lala:~# strace -s 99 -ffp 5796
Process 5796 attached with 3 threads
[pid 5796] futex(0x7f3282ab79d0, FUTEX_WAIT, 36, NULL <unfinished ...>
[pid 5802] restart_syscall(<... resuming interrupted call ...>) = 0
[pid 5802] nanosleep({5, 0}, NULL) = 0
[pid 5802] nanosleep({5, 0}, NULL) = 0
[pid 5802] nanosleep({5, 0}, NULL) = 0
[pid 5802] nanosleep({5, 0}, NULL) = 0
[pid 5802] nanosleep({5, 0}, NULL) = 0
[pid 5802] nanosleep({5, 0}, ^CProcess 5796 detached
Process 5800 detached
Process 5802 detached
<detached ...>
Daniel Nguyen
W dniu 22.07.2017 o 01:43, Mike Jumper [via Apache Guacamole
(incubating) - Users] pisze:
> On Fri, Jul 21, 2017 at 2:16 PM, danielos <[hidden email]
> </user/SendEmail.jtp?type=node&node=1368&i=0>> wrote:
>
>
> Actually I found something (whole guacd log attached to e-mail
> message):
>
> guacd[1]: INFO: Connection "$5ac08735-d767-419f-997a-a80e74f6704b"
> removed.
> guacd[1]: INFO: Creating new client for protocol "rdp"
> guacd[1]: INFO: Connection ID is
> "$7293ee8c-133f-4844-af8f-ce124b9693e3"
> guacd[289]: INFO: No security mode specified. Defaulting to RDP.
> guacd[289]: INFO: Resize method: none
> guacd[289]: INFO: User
> "@09c1c2a7-7554-4453-85a1-19dda4c8cd96" joined connection
> "$7293ee8c-133f-4844-af8f-ce124b9693e3" (1 users now present)
> guacd[289]: ERROR: Password authentication failed:
> Authentication failed (username/password)
> guacd[289]: INFO: User
> "@09c1c2a7-7554-4453-85a1-19dda4c8cd96" disconnected (0 users remain)
> guacd[289]: INFO: Last user of connection
> "$7293ee8c-133f-4844-af8f-ce124b9693e3" disconnected
> *** Error in `/usr/local/sbin/guacd': free(): invalid pointer:
> 0x00007fb74c00b5a0 ***
> ======= Backtrace: =========
> /lib64/libc.so.6(+0x7c503)[0x7fb763e51503]
>
> /usr/local/lib/libguac-client-rdp.so(guac_common_ssh_destroy_user+0x23)[0x7fb75c4117d3]
>
> /usr/local/lib/libguac-client-rdp.so(guac_rdp_client_free_handler+0x61)[0x7fb75c405ba1]
> /usr/local/lib/libguac.so.12(guac_client_free+0x32)[0x7fb7658bfb72]
> /usr/local/sbin/guacd[0x404310]
> /usr/local/sbin/guacd[0x403a80]
> /lib64/libpthread.so.0(+0x7dc5)[0x7fb764efddc5]
> /lib64/libc.so.6(clone+0x6d)[0x7fb763ecc73d]
> ======= Memory map: ========
> 00400000-00407000 r-xp 00000000 00:26 23 /usr/local/sbin/guacd
> 00606000-00607000 r--p 00006000 00:26 23 /usr/local/sbin/guacd
> 00607000-00608000 rw-p 00007000 00:26 23 /usr/local/sbin/guacd
> 00d6b000-0158a000 rw-p 00000000 00:00 0 [heap]
>
>
> That particular issue is:
>
> https://issues.apache.org/jira/browse/GUACAMOLE-194
>
> which should be fixed on the current 0.9.13-incubating release
> candidate. I don't think this would result in guacd spinning and
> eating 100% CPU, but it's worth trying the "0.9.13-incubating-RC1" tag
> of the Docker images to see if the issue remains reproducible.
>
> Are you able to reproduce this outside of Docker?
>
> - Mike
>
>
>
> ------------------------------------------------------------------------
> If you reply to this email, your message will be added to the
> discussion below:
> http://apache-guacamole-incubating-users.2363388.n4.nabble.com/guacd-0-9-12-incubating-is-using-100-cpu-tp1365p1368.html
>
>
> To unsubscribe from guacd 0.9.12-incubating is using 100% cpu, click
> here
> <http://apache-guacamole-incubating-users.2363388.n4.nabble.com/template/NamlServlet.jtp?macro=unsubscribe_by_code&node=1365&code=Z3VhY2Ftb2xlQGRhbi10ZWNoLnBsfDEzNjV8OTM3MTEyMDU2>.
> NAML
> <http://apache-guacamole-incubating-users.2363388.n4.nabble.com/template/NamlServlet.jtp?macro=macro_viewer&id=instant_html%21nabble%3Aemail.naml&base=nabble.naml.namespaces.BasicNamespace-nabble.view.web.template.NabbleNamespace-nabble.view.web.template.NodeNamespace&breadcrumbs=notify_subscribers%21nabble%3Aemail.naml-instant_emails%21nabble%3Aemail.naml-send_instant_email%21nabble%3Aemail.naml>
>
>
--
View this message in context:
http://apache-guacamole-incubating-users.2363388.n4.nabble.com/guacd-0-9-12-incubating-is-using-100-cpu-tp1365p1369.html
Sent from the Apache Guacamole (incubating) - Users mailing list archive at
Nabble.com.
