Hello,

Just installed a new 8 core machine with Fedora Core. I tried to add it to the hosts in BoincView (1.2.5 on a Windows XP machine).

But it does not get any response back from the new machine. The output from the new machine says:

04-Mar-2013 17:40:01 [---] Starting BOINC client version 7.0.28 for x86_64-pc-linux-gnu
04-Mar-2013 17:40:01 [---] log flags: file_xfer, sched_ops, task, cpu_sched, cpu_sched_debug, file_xfer_debug
04-Mar-2013 17:40:01 [---] log flags: http_debug, http_xfer_debug
04-Mar-2013 17:40:01 [---] Libraries: libcurl/7.21.7 NSS/3.12.10.0 zlib/1.2.5 libidn/1.22 libssh2/1.2.7
04-Mar-2013 17:40:01 [---] Data directory: /home/dahls/BOINC
04-Mar-2013 17:40:01 [---] Processor: 8 AuthenticAMD AMD FX(tm)-8120 Eight-Core Processor [Family 21 Model 1 Stepping 2]
04-Mar-2013 17:40:01 [---] Processor: 2.00 MB cache
04-Mar-2013 17:40:01 [---] Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc extd_apicid aperfmperf pni pclmulqdq monitor ssse3 cx16 sse4_1 sse4_2 popcnt aes xsave avx lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs xop skinit wdt lwp fma4 nodeid_msr topoext perfctr_core arat cpb npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold
04-Mar-2013 17:40:01 [---] OS: Linux: 3.1.0-7.fc16.x86_64
04-Mar-2013 17:40:01 [---] Memory: 31.37 GB physical, 15.62 GB virtual
04-Mar-2013 17:40:01 [---] Disk: 781.66 GB total, 734.97 GB free
04-Mar-2013 17:40:01 [---] Local time is UTC +1 hours
04-Mar-2013 17:40:01 [---] No usable GPUs found
04-Mar-2013 17:40:01 [---] Config: GUI RPC allowed from any host
04-Mar-2013 17:40:01 [---] Config: GUI RPC allowed from:
04-Mar-2013 17:40:01 [---] Config:   192.168.1.10
04-Mar-2013 17:40:01 [SETI@home] URL http://setiathome.berkeley.edu/; Computer ID 6923550; resource share 100
04-Mar-2013 17:40:01 [SETI@home] General prefs: from SETI@home (last modified 03-Jul-2011 09:48:53)
04-Mar-2013 17:40:01 [SETI@home] Host location: none
04-Mar-2013 17:40:01 [SETI@home] General prefs: using your defaults
04-Mar-2013 17:40:01 [---] Preferences:
04-Mar-2013 17:40:01 [---]    max memory usage when active: 16063.53MB
04-Mar-2013 17:40:01 [---]    max memory usage when idle: 28914.35MB
04-Mar-2013 17:40:01 [---]    max disk usage: 100.00GB
04-Mar-2013 17:40:01 [---]    don't use GPU while active
04-Mar-2013 17:40:01 [---]    (to change preferences, visit the web site of an attached project, or select Preferences in the Manager)
04-Mar-2013 17:40:01 [---] [cpu_sched_debug] Request CPU reschedule: Prefs update
04-Mar-2013 17:40:01 [---] [cpu_sched_debug] Request CPU reschedule: Startup
04-Mar-2013 17:40:01 [---] Not using a proxy

I have tried to change the password (in the gui_rpc_auth.cfg file), but no success. There is a firewall between, but I have checked it and it does not drop/block any traffic (it's the same rule in the fw that other hosts are using).

I have tried both 31416 and 1043 but that does not hep.

Any suggestions what I can do to make this work?

---

No immediate suggestions, just encouragement to keep trying - it should work. I run a similar configuration, with minor differences:

BoincView, again running on Windows XP, is v1.4.2 - but I don't remember any problems running the earlier versions.

Remote machines are also Windows - ranging from Win 2K Server to Win7/64, and running BOINCs from 6.10.58 to 7.0.52: all are monitored properly.

Your Windows monitoring machine is on IP 192.168.1.10, I take it? And both machines are on the same 192.168.1.0 subnet?

The password in gui_rpc_auth.cfg is certainly important, and the one to use in BoincView. If you change it, you probably need to restart the remote BOINC before you can connect with the local viewer.

If your local Windows machine is running BOINC, can you (for test purposes) connect to the remote computer using 'Select computer...' in BOINC Manager? (top line in Advanced menu)

I have never had any problems with the other linux hosts which the windows machine is connecting to.

Yes, the 192.168.1.10 is the IP of the window machine. No, the other host (I got 5 other linux hosts and 2 other window hosts connected) is on another network and the traffic has to go through a firewall. But all the other hosts are also behind the same firewall (and the same rule in the firewall is used to allow this traffic).

The window machine is also running boinc. But that one I cannot connect through network but by file.

I restarted boinc after I changed the config files.

---

BTW, it's version boinc_7.0.28_x86_64-pc-linux-gnu.sh I'm running the the new linux host.

---

For testing you could trying using boinccmd between the two machines. Something simple like:
boinccmd --host 192.168.101.10 --passwd password --get_host_info
should be good.

The start info shows you are using the cc_config.xml option and the remote_hosts.cfg option to define access.
04-Mar-2013 17:40:01 [---] Config: GUI RPC allowed from any host
04-Mar-2013 17:40:01 [---] Config: GUI RPC allowed from:
04-Mar-2013 17:40:01 [---] Config: 192.168.1.10
Maybe this is confusing the client?

That host doesn't have any kind of SW firewall running be default does it?

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Hm...
I tried this from the windows machine: "can't connect to..."

Then I tried it from a linux host at the same network as the "bad" server: "./boinccmd: error while loading shared libraries: libssl.so.0.9.8: cannot open shared object file: No such file or directory"

Then I tried it on the "bad" host itself: "Authorization failure: -155".

I have tried with and without the remote_hosts.cfg without any luck.

---

Hm...
I tried this from the windows machine: "can't connect to..."

Then I tried it from a linux host at the same network as the "bad" server: "./boinccmd: error while loading shared libraries: libssl.so.0.9.8: cannot open shared object file: No such file or directory"

Then I tried it on the "bad" host itself: "Authorization failure: -155".

I have tried with and without the remote_hosts.cfg without any luck.

That does seem to show port 31416 isn't open between the windows & Linux host. Otherwise you would have gotten the "Authorization failure" message from there as well.

Is that host command line only or does it have a GUI environment installed? If there is a GUI environment. Does BOINC Manager run & connect to the client correctly? If it doesn't I would suspect gui_rpc_auth.cfg is somehow damaged or unreadable by the client apps. Do they run under an account that can not read that file?

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

On my raspberry Pi which runs Debian I had to add the monitoring host into /etc/hosts before I could connect. See my blog as I have instructions for setting up BT talking to the Pi.

---

BOINC blog

Hm...
I tried this from the windows machine: "can't connect to..."

Then I tried it from a linux host at the same network as the "bad" server: "./boinccmd: error while loading shared libraries: libssl.so.0.9.8: cannot open shared object file: No such file or directory"

Then I tried it on the "bad" host itself: "Authorization failure: -155".

I have tried with and without the remote_hosts.cfg without any luck.

That does seem to show port 31416 isn't open between the windows & Linux host. Otherwise you would have gotten the "Authorization failure" message from there as well.

Is that host command line only or does it have a GUI environment installed? If there is a GUI environment. Does BOINC Manager run & connect to the client correctly? If it doesn't I would suspect gui_rpc_auth.cfg is somehow damaged or unreadable by the client apps. Do they run under an account that can not read that file?

The port is open, otherwise I would not have contact with the other hosts. And the firewall log show that the connection attempt is accepted.

It's only the window machine that has the GUI environment. The other host don't even have a monitor attached, so no need for GUI environment.

The installation and the client is running under my regular user, as on all the other hosts.

---

I did a tcpdump on the firewall. The result confuse me a bit:

20:41:04.944461 192.168.1.10.4588 > 192.168.2.21.1043: S 3738152998:3738152998(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:04.944647 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:07.897837 192.168.1.10.4588 > 192.168.2.21.1043: S 3738152998:3738152998(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:07.897988 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:10.429361 192.168.1.10.4589 > 192.168.2.21.1043: S 643657131:643657131(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:10.429511 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:10.945342 192.168.1.10.4590 > 192.168.2.21.31416: S 56035063:56035063(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:10.945499 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:13.430447 192.168.1.10.4589 > 192.168.2.21.1043: S 643657131:643657131(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:13.430589 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:13.933190 192.168.1.10.4590 > 192.168.2.21.31416: S 56035063:56035063(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:13.933352 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:16.429499 192.168.1.10.4592 > 192.168.2.21.31416: S 3002114644:3002114644(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:16.429659 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:19.365125 192.168.1.10.4592 > 192.168.2.21.31416: S 3002114644:3002114644(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:19.365274 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]
20:41:25.989791 192.168.1.10.4594 > 192.168.2.21.1043: S 2683071891:2683071891(0) win 65535 <mss 1460,nop,nop,sackOK> (DF)
20:41:25.989938 192.168.2.21 > 192.168.1.10: icmp: host 192.168.2.21 unreachable - admin prohibited [tos 0xc0]

I'm not sure which host is trying a icmp request, but it seem to be stopped. Does the new boinc client require icmp traffic too? There is no rules that allow the other hosts icmp access through the firewall.

---

About the icmp problem. It was a firewall installed on the new linux host. I disabled it, but that did not help. The only visible change in BoincView was that the ?-mark on the pc icon in the "Locations" list changed to a red X.

---

About the icmp problem. It was a firewall installed on the new linux host. I disabled it, but that did not help. The only visible change in BoincView was that the ?-mark on the pc icon in the "Locations" list changed to a red X.

On my BV setup, the ?-mark icon on a host tends to mean that the BOINC client has crashed, or a loose network cable has fallen out of my switch. Either way, the host is fundamentally unreachable.

The red-X icon tends to mean that the host is (temporarily) too busy to respond to BV's request for information - I usually see it when a memory-intensive application like Einstein or CPDN is just starting a new task.

I don't know what the official definitions of the icons are - I don't think it was ever documented, and of course the online help from the BV website is long gone.

After I came back from Easter vacation I started up boinc view again. Now it works!! Maybe the new host was a bit shy... ;)

---