Jump to content
Sign in to follow this  
jsauve

RedHat Enterprise Linux 3.0 and Linux Client

Recommended Posts

So when is Dantz going to support Redhat Enterprise edition? After all, RH has officially cancelled their product line and support for all versions up to 8.0. Version 9.0 is officially terminated as of April 31, 2004.

 

I've tried using the 6.5 linux client on RH EL 3.0 but it fails (understandably so if Dantz doesn't support it).

 

You should either consider dropping RedHat client support alltogether or keep up with the times and add support for the only remaining RedHat products left (the Enterprise line), this would seem the natural thing to do for an if Retrospect is Enterprise backup solution.

Share this post


Link to post
Share on other sites

Strange. The 6.5 client works fine for me on RHEL 3. (For that matter, the 6.0 client seemed to work as well before I upgraded to 6.5.) What kind of problem are you experiencing?

Share this post


Link to post
Share on other sites

The child process hangs at random and Retrospect can no longer connect to the client until you kill each parent/child process and restart it manually.

 

The following shows up in the syslog

 

Mar 10 00:42:52 doom Retrospect[31592]: connTcpConnection: invalid code found: 111

Mar 10 23:24:51 doom Retrospect[31592]: SThreadSpawn: pthread_create() failed with error 12

Mar 11 09:58:51 doom Retrospect[31592]: SThreadSpawn: pthread_create() failed with error 12

Share this post


Link to post
Share on other sites
Guest psykoyiko

Strange...

 

error 12 (from /usr/include/asm-generic/errno-base.h) seems to indicate ENOMEM - out of memory. I would be curious to see what top/free show when the client is in this condition.

Share this post


Link to post
Share on other sites

Have you found a solution for this?

 

I have about the same problem. Same first error message

 

connTcpConnection: invalid code found: 111

 

but not the following SThreadSpwan... lines.

 

After a while a have a number of defunct client processes that I need to kill manually to get it working again.

 

I do get other messages along with these above saying

 

iplud: duplicate A/C with: Foo_client (192.168.0.xxx:61697)

 

I posted this particular issue also in the Server, SBS and Multi Server forum (post #39629) since I somehow believe that they are not connected (I may be wrong).

Share this post


Link to post
Share on other sites

Unfortunately ... I haven't had any success. But if you check /var/log/retroclient.log, you will see something like this ...

 

 

 

1080851428: Connection established by 198.20.xxx.xxx:2182

 

1080851792: connTcpConnection: invalid code found: 111

 

1080851792: ConnReadData: Connection with 198.20.xxx.xxx:2182 closed

 

1080876905: Connection established by 198.20.xxx.xxx:2583

 

1080876939: connTcpConnection: invalid code found: 111

 

1080876939: ConnReadData: Connection with 198.20.xxx.xxx:2583 closed

 

1080915532: Connection established by 198.20.xxx.xxx:3253

 

1080915554: SThreadSpawn: pthread_create() failed with error 12

 

1080963221: SThreadSpawn: pthread_create() failed with error 12

Share this post


Link to post
Share on other sites

I have one machine running Red Hat Linux 9.0, and the RetroClient.log reads as follows:

 

1085417659: ConnStartListen: starting thread ConnStartListen for 127.0.0.1:0

1085417659: iplud: bound to address 0.0.0.0

1085417659: ipludAddMembership: adding membership for 0.0.0.0

1085417665: IPNSRegister(0): registered: "SERVER-CVS (linux CVS)"/"b96258547e1773ad"

1085417665: ConnStartListen: starting thread ConnStartListen for 10.1.2.54:0

1085417671: IPNSRegister(0): registered: "SERVER-CVS (linux CVS)"/"b96258547e1773ad"

1085421629: Connection established by 10.1.2.53:1035

1085421629: ConnReadData: Connection with 10.1.2.53:1035 closed

1085421629: ServicePurge: service not found

This machine is visible from my Retrospect Multi Server 6.0 (or 6.5).

 

I have another machine running Red Hat Enterprise Linux 3.0, and the RetroClient.log reads as follows:

 

1085159322: ConnStartListen: starting thread ConnStartListen for 127.0.0.1:0

1085159322: iplud: bound to address 0.0.0.0

1085159322: ipludAddMembership: adding membership for 0.0.0.0

1085159328: IPNSRegister(1): registered: "IGstk"/"93c5a93c5af3c995"

1085159328: ConnStartListen: starting thread ConnStartListen for 10.1.2.56:0

1085159334: IPNSRegister(1): registered: "IGstk"/"17aa550c004cc631"

 

This machine is not visible by the Retrospect Multi Server 6.0 (or 6.5).

 

Can any one figure out what the problem is? What is this IPNSRegister with parameter 1.

I have only one ethernet card on the machine.

 

 

Sohan

~

~ computer.gif

Share this post


Link to post
Share on other sites

I am still not clear if Retrospect is going to support Red Hat Enterprise Linux

or not? Even a customer support guy (and we have a support contract with

Dantz) just reiterated that "we support only till Red Hat Linux 8.0", but we

have heard from people that the Linux client has also worked on Enterprise

Linux". He didn't even listen to the error messages I get when I try using,

which also appears on this page.

Share this post


Link to post
Share on other sites

Hi

 

Dantz has not officially tested anything other than RedHat 7.X,8.X and 9 for compatibility with the client application. As a result other versions of RedHat or linux distributions are "officially" unsupported.

 

We are aware that these versions are out of date and will address the issue in future versions of Retrospect.

 

There are a lot of users who are using the client with different Linux distributions and RedHat versions. While unsupported by Dantz it does appear to work.

 

Thanks

Nate

Share this post


Link to post
Share on other sites

If so, then can someone please see message #41979 - 05/27/04 10:49 AM

above, and suggest what possible error could be, assuming that the OS is

Retrospect supported Linux?

Share this post


Link to post
Share on other sites

Hi

 

have you tried binding it to the address of the NIC using retroclient --ip? What version of the client do you have installed?

 

Thanks

Nate

Share this post


Link to post
Share on other sites

I tried that a long time ago by modifying the rcl startup script as you see below. As far as I'm concerned, Dantz has a bad reputation now, this issue should have been addressed ages ago, it has been long known that RH 7.x-9.0 was going to be dropped (years, not months ago). Compounded on top of this issue, the support contract on a "Per-module" basis system is ludacris to put it as nice as I can. May I also add that the issue of the Retrospect Helper Service issue on Windows should have been fixed years ago too. I'm still experiencing that problem and it was first addressed two years ago by users in the forums. No point in going on with my list of grievances because it gets nowhere with Dantz.

 

# Retrospect Client rcl script

#

# Contact Information

#

# Dantz Development Corporation

# 4 Orinda Way, Building C

# Orinda, CA 94563

# 800.225.4880 or 925.253.3000

# customer_service@dantz.com

#

# Copyright 2000-2002 Dantz Development Corporation

#

#

# Place this script along with retroclient, and

# retrocpl in the directory specified by CLIENTDIR.

#

CLIENTDIR=/usr/local/dantz/client

case $1 in

start)

$CLIENTDIR/retroclient -ip 198.20.xxx.xxx -daemon

;;

stop)

$CLIENTDIR/retrocpl -stop

;;

status)

$CLIENTDIR/retrocpl

;;

*)

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

;;

esac

exit 0

Share this post


Link to post
Share on other sites

available in "future versions" ... ? I suppose that means a non-cost-free upgrade to Retrospect 7.0, since thats how the upgrades seem to go with Dantz. This is something that should have worked in 6.5, heck, even 6.0.

Share this post


Link to post
Share on other sites
Guest psykoyiko

JasonS-

 

It looks to me like the client is running out of memory -- error 12 is defined as ENOMEM on x86. Can you post the output of top -n 1 and free when the client is in this condition?

 

Thanks.

Share this post


Link to post
Share on other sites

I've ran a backup of the server as of today and it went smoothly. This may be due to Redhat Enterprise 3 Update 2 being installed. Previous conditions where the error occurred were with Enterprise 3, before any of the quarterly updates were rolled out.

 

I will continue monitoring and provide any diagnostic output should the situation arise again.

Share this post


Link to post
Share on other sites

I've noticed that adding -ip x.x.x.x in the /etc/rc.d/init.d/rcl startup script doesn't make retroclient listen on only the specified IP address. I've reproduced this on RH7.3 and RHEL 3.0 Update 2.

 

On both systems if I do a netstat -an | grep ':497 ', it shows the client daemon listening on all interfaces

 

tcp 0 0 198.20.x.x:497 0.0.0.0:* LISTEN

tcp 0 0 127.0.0.1:497 0.0.0.0:* LISTEN

udp 0 0 0.0.0.0:497 0.0.0.0:*

Share this post


Link to post
Share on other sites

Similarily, `lsof | grep retro` confirms the netstat results

 

retroclie 18184 root 0u IPv4 4198106 TCP localhost.localdomain:497 (LISTEN)

retroclie 18184 root 1u IPv4 4198618 TCP xxx.robarts.ca:497 (LISTEN)

retroclie 18184 root 2u IPv4 4198108 UDP *:497

Share this post


Link to post
Share on other sites

/var/log/retroclient.log contains the following messages still. The client still hasn't hung yet. I suspect that a kernel update in RHEL 3 Update 2 may have fixed the previous problems I had been experiencing with the error 12 ENOMEM situation.

 

1091124604: ServicePurge: service not found

1091159309: connTcpConnection: invalid code found: 111

 

Here's some additional information that I've found in /var/log/messages produced by retrospect

 

Jul 29 14:05:49 smack Retrospect[5780]: iplud: bound to address 0.0.0.0

Jul 29 14:05:49 smack Retrospect[5780]: ipludAddMembership: adding membership for 0.0.0.0

Jul 29 14:05:55 smack Retrospect[5780]: IPNSRegister(1): registered: "smack"/"e79c47639b776254"

Jul 29 14:06:01 smack Retrospect[5780]: IPNSRegister(1): registered: "smack"/"782121639b776254"

Jul 29 14:06:32 smack Retrospect[5780]: IPNSRegister(0): registered: "smack"/"782121639b776254"

Jul 29 14:06:32 smack kernel: application bug: retroclient(5808) has SIGCHLD set to SIG_IGN but calls wait().

Jul 29 14:07:48 smack kernel: application bug: retroclient(5906) has SIGCHLD set to SIG_IGN but calls wait().

Jul 29 14:10:04 smack Retrospect[5780]: ServicePurge: service not found

Jul 29 14:10:06 smack kernel: application bug: retroclient(6036) has SIGCHLD set to SIG_IGN but calls wait().

Jul 29 23:48:29 smack Retrospect[5780]: connTcpConnection: invalid code found: 111

Share this post


Link to post
Share on other sites

Retrospect client died yet again with the following in /var/log/retroclient.log

 

1091541832: SThreadSpawn: pthread_create() failed with error 12

1091541872: SThreadSpawn: pthread_create() failed with error 12

 

I've attached the output of free and `top -n 1` at the time of death.

 

10:04:52 up 6 days, 9:27, 3 users, load average: 4.45, 5.14, 2.60

97 processes: 96 sleeping, 1 running, 0 zombie, 0 stopped

CPU states: cpu user nice system irq softirq iowait idle

total 0.0% 0.0% 0.2% 0.0% 0.0% 2.7% 97.0%

cpu00 0.0% 0.0% 0.9% 0.0% 0.0% 3.9% 95.0%

cpu01 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 100.0%

cpu02 0.0% 0.0% 0.0% 0.0% 0.0% 6.9% 93.0%

cpu03 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 100.0%

Mem: 511916k av, 495468k used, 16448k free, 0k shrd, 7412k buff

321144k actv, 59900k in_d, 6808k in_c

Swap: 883564k av, 147048k used, 736516k free 320436k cached

 

PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND

18134 root 20 0 1112 1112 772 R 0.2 0.2 0:00 0 top

1 root 15 0 504 472 448 S 0.0 0.0 0:12 3 init

2 root RT 0 0 0 0 SW 0.0 0.0 0:00 0 migration/0

3 root RT 0 0 0 0 SW 0.0 0.0 0:00 1 migration/1

4 root RT 0 0 0 0 SW 0.0 0.0 0:00 2 migration/2

5 root RT 0 0 0 0 SW 0.0 0.0 0:00 3 migration/3

6 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 keventd

7 root 34 19 0 0 0 SWN 0.0 0.0 0:00 0 ksoftirqd/0

8 root 34 19 0 0 0 SWN 0.0 0.0 0:00 1 ksoftirqd/1

9 root 34 19 0 0 0 SWN 0.0 0.0 0:00 2 ksoftirqd/2

10 root 34 19 0 0 0 SWN 0.0 0.0 0:00 3 ksoftirqd/3

13 root 15 0 0 0 0 SW 0.0 0.0 0:00 2 bdflush

11 root 15 0 0 0 0 SW 0.0 0.0 0:48 0 kswapd

12 root 15 0 0 0 0 SW 0.0 0.0 0:14 0 kscand

14 root 15 0 0 0 0 SW 0.0 0.0 0:22 1 kupdated

15 root 25 0 0 0 0 SW 0.0 0.0 0:00 2 mdrecoveryd

21 root 25 0 0 0 0 SW 0.0 0.0 0:00 3 scsi_eh_0

24 root 15 0 0 0 0 SW 0.0 0.0 3:21 2 kjournald

75 root 25 0 0 0 0 SW 0.0 0.0 0:00 1 khubd

568 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 kjournald

2139 root 15 0 588 564 508 S 0.0 0.1 1:28 3 syslogd

2143 root 15 0 456 396 392 S 0.0 0.0 0:00 1 klogd

2153 root 15 0 452 444 388 S 0.0 0.0 0:30 1 irqbalance

2161 rpc 15 0 540 464 460 S 0.0 0.0 0:00 0 portmap

2180 rpcuser 15 0 1008 620 616 S 0.0 0.1 0:00 2 rpc.statd

2193 root RT 0 560 456 424 S 0.0 0.0 0:16 0 auditd

2276 nscd 15 0 2984 2160 856 S 0.0 0.4 3:02 3 nscd

2327 root 15 0 860 692 608 S 0.0 0.1 0:00 1 sshd

2341 root 15 0 776 740 656 S 0.0 0.1 0:24 0 xinetd

2354 ntp 15 0 2564 2564 2192 S 0.0 0.5 0:21 2 ntpd

2378 root 25 0 544 412 408 S 0.0 0.0 0:00 1 safe_mysqld

2409 mysql 15 0 1112 336 332 S 0.0 0.0 0:00 1 mysqld

 

-(root@doom)-(/var/log)--

# free

total used free shared buffers cached

Mem: 511916 495176 16740 0 7412 320444

-/+ buffers/cache: 167320 344596

Swap: 883564 147048 736516

Share this post


Link to post
Share on other sites

Jason S, did you ever get any answers to this? I just tracked down a crashing bug that I've been seeing to this, as well -- I have a client running on RedHat, and it dies halfway through the backup. Digging through the logfiles, every time it dies, it's preceded by one second by a line:

 

SThreadSpawn: pthread_create() failed with error 12

 

I have a gig of RAM in the client machine, and it continues doing everything else it's doing at that time without a problem, but unfortunately I haven't been sitting next to it able to check out the memory situation right when it crashes.

 

I'd love to get a solution to this, as it makes backing up the machine nigh impossible.

Share this post


Link to post
Share on other sites

Hi Jason L,

 

Unfortunately we have had nothing further to report on this bug. We've had so many issues dealing with dantz, and the product itself where it has clearly been a product issue and not an "end-user problem" that we've lost all faith in them. They make obtaining free support when you have a product bug anything, except possible.

 

All of our issues to date (at least 4) have been regarding software bugs. They refuse to support their own coding flaws without taking a credit card and charging ~ $100 USD per incident, heck, they won't even talk to you regarding them until you give them payment.

 

Good luck getting a refund after you prove to them its a bug in their software, it took me a month to obtain one even after they admitted it was a bug only 2 days after I reported the issue (SQL agent addon bug - would cause retrospect to hang indefinately). What do you know, still no resolution even after reporting that bug and having their "engineers" confirm it as such.

 

Its obvious that nobody there reads the support forums either. Their documentation is to say the least, poor with too much ambiguity or too little clarity. Retrospect documentation often contradicts what the application does and we often have to find out what things do by "trial and error" to see that they don't operate as documented.

 

I've had much superior support from Sony (gasp) regarding our tape libraries than Dantz.

 

We've ceased all spending on Dantz products. We don't agree with their support model - purchasing support on a per-dantz-module yearly basis is ludacris.

Share this post


Link to post
Share on other sites

Hello Jason,

 

Have you been able to reproduce the failure on a supported client OS?

I certainly understand your frustration but Dantz has never claimed to support RH enterprise linux. Users are welcome to try of course but there is always the possibility it just wont work. In the current situation there isn't much Dantz can do for you. Once your OS is supported you will find the response to be much different.

 

Thanks

Nate

Share this post


Link to post
Share on other sites

Natew, I'm not running it on Enterprise Linux 3.0, I'm running it on RedHat 9, which you say above is officially supported. That being said, I'm seeing the same exact bug, and the lack of any solution has made me start to price out other backup options.

Share this post


Link to post
Share on other sites

Hi Jason,

 

Reading your posts it seems very clear you are using both RH enterprize linux and other supported OSes. The title of the thread also gives that impression. From a support perspective we can't do a thing about the non supported OSes. If we are going to investigate this we need to focus _only_ on the supported OSes.

 

If you would like to start a thread dealing exclusively with a RH9 client I will be happy to investigate this further. Since the issue appears to be the same on all of your computers lets focus on one machine only. Please post system details and relevant logs. I'll do what I can to take it from there.

 

Thanks

Nate

Share this post


Link to post
Share on other sites

Nate:

 

"My posts"?!? I have exactly two posts in this thread: one dated 10/12/04 03:34 PM, and one dated 10/27/04 05:24 AM, both authored by me ("Jason Levine"). There are a lot by "Jason S", who is NOT me (a fact that I would assume you can easily check by looking at the accounts used to log in and post the messages). I am running RedHat 9, period; I posted in this thread because it was the thread that spelled out, exactly, what the bug was that I was seeing and that was completely killing my ability to use Retrospect to back up my machine.

 

So do you still want me to create another thread? If so, let me know and I'll post details there; otherwise, let me know that too and I'll post them here.

 

/jason

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×