Jump to content

Linux Client RHEL 4


davidfsmith

Recommended Posts

Right I've got RHEL 4 installed on a machine with dual NIC's and it is running as:

 

/usr/local/dantz/client/retroclient -ip 192.168.10.18 -daemon

 

So I could try and debug an issue with it not showing on my Mac Backup server, however when I tried to change the log level using retrocpl:

 

/usr/local/dantz/client/retrocpl -log 10

 

I got:

 

1154688334: pmcWriteData: send() failed with error 88

DoStatus: PmcGetStatus(1) failed with error 519

[root@digital david]# /usr/local/dantz/client/retrocpl

1154688337: pmcWriteData: send() failed with error 88

DoStatus: PmcGetStatus(1) failed with error 519

 

The log:

 

tail -f /var/log/retroclient.log

 

1154688332: ConnStartListen: starting thread ConnStartListen for 127.0.0.1:0

1154688332: iplud: bound to address 0.0.0.0

1154688332: ipludAddMembership: adding membership for 0.0.0.0

1154688338: IPNSRegister(1): registered: "xxxxxxx.xx.xxxxxxxxxxxxxxxxx.xxx1cefa5e1c915beb7"/"1cefa5e1c915beb7"

1154688338: ConnStartListen: starting thread ConnStartListen for 192.168.10.18:0

1154688344: IPNSRegister(1): registered: "xxxxxxx.xx.xxxxxxxxxxxxxxxxx.xxx1cefa5e1db74962a"/"1cefa5e1db74962a"

 

I've got 2 other servers running RHEL4 one is the exact same spec, and retrospect works fine, this is a clean install, and fully updated....

 

I've also gone through other posts on the forum here to try and see if I can resolve the issue, I've commented out RETROSPECT_HOME in /etc/profile as detailed elsewhere.

 

iptables is shutdown

 

If I try and add the client it shows as "Not Activated" and doesn't allow me to select "Configure" to add it, if I try and add by IP Address the "Not Activated" changes to "Responding" and then back to "Not Activated" and the log has:

 

1154689217: Connection established by 192.168.10.20:49382

1154689217: ConnReadData: Connection with 192.168.10.20:49382 was reset

1154689217: ServicePurge: service not found

1154689246: Connection established by 192.168.10.20:49383

1154689246: ConnReadData: Connection with 192.168.10.20:49383 was reset

1154689246: ServicePurge: service not found

1154689246: Connection established by 192.168.10.20:49384

1154689246: ConnReadData: Connection with 192.168.10.20:49384 was reset

1154689246: ServicePurge: service not found

1154689246: Connection established by 192.168.10.20:49385

1154689246: ConnReadData: Connection with 192.168.10.20:49385 was reset

1154689246: ServicePurge: service not found

 

On the server I'm trying to add.

 

Any help, thoughts, much appreciated on this.

 

(and yes the client is the latest version:

rpm -q retroclient

retroclient-7.0.110-1

 

Regards

Link to comment
Share on other sites

hi BN,

 

when you say you have the client started like this:

 

Quote:

/usr/local/dantz/client/retroclient -ip 192.168.10.18 -daemon

 

 


 

how exactly are you doing that? i'm running a similar config on RHES 3 but i modified the rcl script like this:

 

Quote:

case $1 in

start)

$CLIENTDIR/retroclient -daemon -ip XXX.XXX.XXX.XXX

;;

stop)

$CLIENTDIR/retrocpl -stop

;;

status)

$CLIENTDIR/retrocpl

;;

*)

echo "Usage: $0 {start|stop|status}"

;;

esac

 

 


 

and also, when you say:

 

Quote:

I've got 2 other servers running RHEL4 one is the exact same spec, and retrospect works fine

 


 

does this other machine that works also have the dual NIC and you are binding the ip?

 

one final note: i don't think upping the logging on Retroclient will help you too much here. what i'd rather see is a portscan of the linux machine.

 

cheers.

Link to comment
Share on other sites

I've been starting the client on the command line, while trying to work out what's going on... (although i get the same problem if i use the rcl init.d script - which incidentally is unable to stop the process)

 

With regards to the other machines I'm not binding the IP as I didn't have to do anything to make it work, installed the RPM, checked it was running, connected from the backup server, all very nice and painless (on 2 out of the 3 machines....) just this final machine being a total PITA

 

The backup server can talk to the RHEL server (sometimes), however when I try and add in the password to configure it I get back that the password is wrong and the log shows:

 

1154689246: Connection established by 192.168.10.20:49383

1154689246: ConnReadData: Connection with 192.168.10.20:49383 was reset

1154689246: ServicePurge: service not found

1154689246: Connection established by 192.168.10.20:49384

1154689246: ConnReadData: Connection with 192.168.10.20:49384 was reset

1154689246: ServicePurge: service not found

1154689246: Connection established by 192.168.10.20:49385

1154689246: ConnReadData: Connection with 192.168.10.20:49385 was reset

1154689246: ServicePurge: service not found

Link to comment
Share on other sites

hi BN,

 

ok, gotcha. couple of more questions:

 

Quote:

i get the same problem if i use the rcl init.d script - which incidentally is unable to stop the process

 


 

wierdness! i definitely don't have this problem on RHES 3. do the rcl commands work on your other 2 RH machines? i just tried, stop, start, status in that order w/ no problems.

 

unfortunately i don't know how to read these logs. i'm wondering if the error could be misleading however. could you try reseting the password on the client, like this:

 

http://kb.dantz.com/article.asp?article=8346&p=2

 

to a really simple word. something like 'password', and then see if you can log in. you can always change this from the server once you've logged in.

Link to comment
Share on other sites

just port 497 for both TCP and UDP. i have a feeling this will be no problem, but it's best to try the basics first.

 

one other thing. do you see the same type of errors when you do *not* try to bind the client? say if you just started it normally. what was it that made you want to bind originally?

Link to comment
Share on other sites

Ok port scan results are that 497 is open for business for TCP but not UDP (this is the same on one of my other servers where the backup works)

 

# nmap -sT 127.0.0.1

PORT STATE SERVICE

22/tcp open ssh

25/tcp open smtp

111/tcp open rpcbind

497/tcp open dantz

631/tcp open ipp

 

# nmap -sU 127.0.0.1

PORT STATE SERVICE

111/udp open|filtered rpcbind

123/udp open|filtered ntp

631/udp open|filtered unknown

802/udp open|filtered unknown

32768/udp open|filtered omad

 

I bound the client to IP as from going through some of the posts here that was a suggested solution, although it seems to have made no difference in this instance, also I killed the client this morning and then restarted as below:

 

# /etc/init.d/rcl start

# /usr/local/dantz/client/retrocpl

1155194482: pmcWriteData: send() failed with error 88

DoStatus: PmcGetStatus(1) failed with error 519

 

# /etc/init.d/rcl stop

1155204322: pmcWriteData: send() failed with error 88

DoStatus: PmcGetStatus(1) failed with error 519

Link to comment
Share on other sites

hi BN,

 

can you portscan from outside the machine? likesay from one of the other Linux machines? i've seen similar problems before with a client that did not have 497 open for UDP, but i'm not convinced this is the problem since you've only scanned the loopback interface.

 

with the client running and no firewall you should have services on both 497 for TCP and UDP.

Link to comment
Share on other sites

From an external machine only TCP is reporting back

 

# nmap -sT -P0 192.168.10.18

PORT STATE SERVICE

22/tcp open ssh

111/tcp open rpcbind

497/tcp open dantz

 

# nmap -sU 192.168.10.18

PORT STATE SERVICE

111/udp open rpcbind

123/udp open ntp

137/udp open netbios-ns

138/udp open netbios-dgm

631/udp open unknown

753/udp open rrh

32768/udp open omad

Link to comment
Share on other sites

hi BN,

 

ok, now i think we are on to something. why is 497 UDP not showing? we know that iptables is off for this machine, as you stated that in an earlier post.

 

i'm thinking that you should netstat the machine but i'm not sure what that will show us. you could try that and let me do a little research. i'll post back later.

Link to comment
Share on other sites

well just in case it helps

 

# netstat

Active Internet connections (w/o servers)

Proto Recv-Q Send-Q Local Address Foreign Address State

tcp 0 0 xxxxxxx.xx.xxxxxxxxx:ssh ::ffff:192.168.10.82:50881 ESTABLISHED

tcp 0 0 xxxxxxx.xx.xxxxxxxxx:ssh ::ffff:192.168.10.82:50833 ESTABLISHED

Active UNIX domain sockets (w/o servers)

Proto RefCnt Flags Type State I-Node Path

unix 2 [ ] DGRAM 6137 @/var/run/hal/hotplug_socket

unix 2 [ ] DGRAM 3192 @udevd

unix 12 [ ] DGRAM 5069 /dev/log

unix 2 [ ] DGRAM 9473

unix 3 [ ] STREAM CONNECTED 9348

unix 3 [ ] STREAM CONNECTED 9347

unix 3 [ ] STREAM CONNECTED 7435

unix 3 [ ] STREAM CONNECTED 7434

unix 2 [ ] DGRAM 6232

unix 3 [ ] STREAM CONNECTED 6136 /var/run/dbus/system_bus_socket

unix 3 [ ] STREAM CONNECTED 6135

unix 3 [ ] STREAM CONNECTED 6116 /var/run/dbus/system_bus_socket

unix 3 [ ] STREAM CONNECTED 6115

unix 2 [ ] DGRAM 6024

unix 3 [ ] STREAM CONNECTED 6008

unix 3 [ ] STREAM CONNECTED 6007

unix 2 [ ] DGRAM 5893

unix 2 [ ] DGRAM 5874

unix 2 [ ] DGRAM 5859

unix 2 [ ] DGRAM 5840

unix 2 [ ] DGRAM 5667

unix 2 [ ] DGRAM 5584

unix 3 [ ] STREAM CONNECTED 5280

unix 3 [ ] STREAM CONNECTED 5279

unix 2 [ ] DGRAM 5150

unix 2 [ ] DGRAM 5077

Link to comment
Share on other sites

hi BN,

 

that's good, but you could do it like this also (sorry i didn't post this): 'netstat -u -t -a'.

 

two questions though, have you tried to add by ip address? or using the 'Test' button to see if Retrospect can connect that way?

 

this will bypass the UDP, and although that doesn't solve out UDP problem, my thinking is that getting this machine backed up is more of a priority right now.

Link to comment
Share on other sites

Quote:

'netstat -u -t -a'.

 


 

# netstat -u -t -a

Active Internet connections (servers and established)

Proto Recv-Q Send-Q Local Address Foreign Address State

tcp 0 0 *:32768 *:* LISTEN

tcp 0 0 *:sunrpc *:* LISTEN

tcp 0 0 xxxxxxxx.xx.xxxxxx:497 *:* LISTEN

tcp 0 0 localhost.localdomain:497 *:* LISTEN

tcp 0 0 localhost.localdomain:ipp *:* LISTEN

tcp 0 0 localhost.localdomain:smtp *:* LISTEN

tcp 0 0 *:ssh *:* LISTEN

tcp 0 960 xxxxxxxx.xx.xxxxxx:ssh ::ffff:192.168.10.111:51436 ESTABLISHED

udp 0 0 *:32768 *:*

udp 0 0 *:819 *:*

udp 0 0 *:sunrpc *:*

udp 0 0 *:ipp *:*

udp 0 0 xxxxxxxx.xx.xxxxxx:ntp *:*

udp 0 0 localhost.localdomain:ntp *:*

udp 0 0 *:ntp *:*

udp 0 0 *:ntp *:*

 

Quote:

two questions though, have you tried to add by ip address? or using the 'Test' button to see if Retrospect can connect that way?

 

this will bypass the UDP, and although that doesn't solve out UDP problem, my thinking is that getting this machine backed up is more of a priority right now.

 


 

Doesn't work either, it prompts for the password and then comes back with a password error. (even when the password is 1234)

Link to comment
Share on other sites

hi BN,

 

this may be a silly question, but can you open up Retrospect and go to 'Special->Licenses' and then check the text at the bottom of that window to make sure you have another license available for a Client?

 

i've been over and over a couple of things, and now i've done a close reading of your original and subsequent posts and one thing sticks out. you said Retrospect says, "Not Activated", for this client, but it should say, "Not Logged In" if you have an available license.

 

i'm taking off for the rest of the day, but i'll try to check back here tonight after my son goes to bed.

 

cheers.

Link to comment
Share on other sites

hi BN,

 

ok, i did do some more research, but i'm coming up a little empty here. i'm thinking we need to start over (not a big deal) but first some explaination:

 

here is a KB article on 'not activated':

 

http://kb.dantz.com/article.asp?article=5341&p=2

 

doesn't explain much, talks about 'renaming' the computer--personally i don't think that's going to do much for us, but you can use the 'hostname' command on RH to check the hostname, and supposedly you can set the hostname with this command (check the man page) but i have never had luck with this. i usually just change /etc/hostname and reboot the machine.

 

the only other 'not activated' type of error i've ever seen was when i forgot to update a client. since we know that you have the latest, we should probably use RPM to uninstall the client, and then check to make sure there is nothing in /usr/local/dantz/client, and nothing to do w/ Retrospect in /var/log or /etc/init.d. once you've verified that, try a reinstall.

 

after a reinstall with all the defaults, let's try the portscan and netstat commands again to see what everything looks like. then let's continue. i'm hoping this is just a corrupt file or something, but you never know.

Link to comment
Share on other sites

By jove the name change might have cracked it....

 

can't be sure as the backup server is backing up everything else, however another server I have with retrospect installed could see the server, and it was showing as "not logged on" will check once the main backup server has done it's work and post up the result.

 

however for the record I edited the system hostname in /etc/sysconfig/network and changed it using the hostname command to save on the reboot (I also removed retroclient using RPM and re-installed)

 

anyone, fingers crossed all is now good... if so, I owe you beers (or drink of your choice) waltr ;-)

Link to comment
Share on other sites

Sorry about the delay in posting back, looks like the server name was one letter too long (was using the FQDN as the server name)

 

Was

xxxxxxx.xx.xxxxxxxxxxxxxxxxx.xxx (x's to protect the guilty!)

 

So just changed to the first part

xxxxxxx

 

and all was good.

 

Also noticed that when I added the server to the backup system when it came to putting in the name the one I wanted (FQDN) was 1 char too long.

 

Thanks for all the help waltr.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...