No udp-sender?
-
Just rebooted.
Apache2 error.log
[Wed Jan 23 11:49:50.285987 2019] [mpm_event:notice] [pid 11779:tid 139833165219008] AH00489: Apache/2.4.25 (Debian) OpenSSL/1.0.2q configured – resuming normal operations
[Wed Jan 23 11:49:50.286042 2019] [core:notice] [pid 11779:tid 139833165219008] AH00094: Command line: ‘/usr/sbin/apache2’
[Wed Jan 23 14:23:47.088383 2019] [mpm_event:notice] [pid 11779:tid 139833165219008] AH00491: caught SIGTERM, shutting down
[Wed Jan 23 14:24:27.879540 2019] [mpm_event:notice] [pid 642:tid 139808269137088] AH00489: Apache/2.4.25 (Debian) OpenSSL/1.0.2q configured – resuming normal operations
[Wed Jan 23 14:24:28.547030 2019] [core:notice] [pid 642:tid 139808269137088] AH00094: Command line: ‘/usr/sbin/apache2’php-7.0.log
[23-Jan-2019 14:23:47] NOTICE: Terminating …
[23-Jan-2019 14:23:47] NOTICE: exiting, bye-bye!
[23-Jan-2019 14:24:32] NOTICE: fpm is running, pid 499
[23-Jan-2019 14:24:32] NOTICE: ready to handle connections
[23-Jan-2019 14:24:32] NOTICE: systemd monitor interval set to 10000msThat’s all
-
@george1421 said in No udp-sender?:
@URfog said in No udp-sender?:
It’s only to hide my IPs, the real IPs are correct
I guess I asked the question wrong.
Are they all the same IP address or different? If they are different why, if your FOG server has only 1 NIC adapter?
Yes, they are different
First one is local internal IP (intranet), last is external IP. Could it be wrong? -
@URfog Well how are you doing this? Are you using a vlan trunk over the single physical nic?
If you have a public IP address (i.e. internet based IP address issued by your ISP) please mask that address with 1.1.1.1 for internal IP addresses please leave them as they are because I need to see how you have it setup. Please use this command to display your network configuration.
ip addr showand post the output here. Internal or non routable IP addresses are not a security concern if I know. -
This is ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000 link/ether 8c:dc:d4:2e:1e:41 brd ff:ff:ff:ff:ff:ff inet 10.11.0.113/16 brd 10.11.255.255 scope global eth0 valid_lft forever preferred_lft forever inet6 fe80::8edc:d4ff:fe2e:1e41/64 scope link valid_lft forever preferred_lft foreverWhere do FOGMulticastManager service gets my external IP from?
Cheers
[Mod Note] Added code block around output for readability -Geo
-
@URfog said in No udp-sender?:
Where do FOGMulticastManager service gets my external IP from?
This is done by calling http://ipinfo.io/ip to get the external IP. Tom added that a while ago, I think it’s used to be able to connect FOG servers that run behind gateways. It’s not particularly for the Multicast service but all FOG services use this interface enumeration. We removed/disabled that part in the latest development version weeks ago as it’s a special use case.
Re-reading most of the topic and especially your initial post I don’t notice much of an issue here. The only statement I see is “I cannot get multitask working”. What exactly does not work? What are your steps to schedule a multicast task? The usual way is create a group, add hosts and schedule a multicast task for this group. I am fairly sure this is working in 1.5.5 as I have tested this not long ago. But there are other ways of multicasting in FOG. Maybe one of the others is not working. Please tell us exactly what you are trying to do.
-
Hi @Sebastian-Roth, it’s been a “long” time that I’ve been using FOG, it’s not new for me, and it was working.
The issue that I notice is that of FOGMulticastManager not loading ant udp-sender service.
I tried creating a Multicast task from- Group
- Named
- To a single workstation.
And I can see that task are created but, in the clients I see
Attempting to check in… failed
Invalid Multicast Session (In line for .)
In MYSQL I see
- In multicastSessionsAssoc table there is one record per client
- In tasks table there is one record per client
- In multicastSessions table there is no record
So, my question is, should it be one record in multicastSessions table?
If I do ps aux | grep udp Should I find any udp-sender proccess?
Thanks
-
@URfog said in No udp-sender?:
Attempting to check in… failed
Invalid Multicast Session (In line for .)Can we get a clear screen shot of this error taken with a mobile phone (unless the entire bit of text on the screen is “Invalid Multicast Session (In line for .)” then it adds no value)? Does it show you the kernel parameters on the error screen?
The context of the error message sometimes is just as important as the actual error itself.
Just to be sure when you go to schedule a multicast deployment, you do have hosts listed in the “Hosts in task” list?
@Sebastian-Roth Its been a while since I worked with Multicasting images, but I just scheduled one and it didn’t ask me for a session name when I submitted the task. Looking at the “Task Name” column the session name appears to be “Multi-Cast Task” Is this expected? (FWIW I’m still running FOG 1.5.4 in production)
-
Hi @george1421, there they go
-
@URfog Thanks for the details! I’ll have a closer look later on today or tomorrow.
@george1421 said:
Its been a while since I worked with Multicasting images, but I just scheduled one and it didn’t ask me for a session name when I submitted the task. Looking at the “Task Name” column the session name appears to be “Multi-Cast Task” Is this expected?
Got me there. I think I never ever used multicast that way. Back at my old working place where we heavily used multicast we’d always schedule it via groups. I’ll check that out as well.
-
@Sebastian-Roth said in No udp-sender?:
@URfog Thanks for the details! I’ll have a closer look later on today or tomorrow.
@george1421Hi all ( @Sebastian-Roth, @george1421), you can leave this problem, I reinstalled the OS and FOG, and with only one machine and one image in my new FOG it worked, I reused my DB backup and it retruned to fail, so the error was in the DB. I have recreated it and now it’s working.
Thank you so much for your efforts.
-
@URfog Oh wow, didn’t expect that! Thanks heaps for letting us know! Great that you figured this out.
-
What exactly did you end up doing to fix this? I am experiencing the exact same issue with my installation.
I"m running Debian 9.9
FOG 1.5.5Thanks!
Cheers!
-
@Joe-Gill See what @URfog said:
I reused my DB backup and it retruned to fail, so the error was in the DB. I have recreated it and now it’s working.
Possibly you are in the same situation. Not sure if we can figure out what exactly is wrong in the DB. If I’d have heaps more time I’d suggest you send me a dump of your database and I’d try to replicate the issue and find a fix. But unfortunately I can’t afford the time.
Re-creating the whole DB is probably not an option for you, right?
Maybe start by cleaning up your DB: https://wiki.fogproject.org/wiki/index.php/Troubleshoot_MySQL#Database_Maintenance_Commands
-
Yeah I would rather not re-build my DB from scratch if at all possible. I have fewer hosts now than when I started with FOG but it’s still a large number. I am in the process of setting up another FOG server to do a fresh install and go from there. I do need to clean some things up in there so it may be the time to do this. I did run through the Troubleshoot MySQL Database Maintenance Commands and didn’t get real far with it. I"ll look at that once more before I bite the bullet on starting over.
I’ll post back to the other thread I started. Hopefully I can figure something out without having to re-do the entire DB…
Thanks for all of your help! I know you guys put in a ton of time with FOG and it’s very appreciated!!
Cheers!
Joe
