Server admin unresponsive, any way to fix it?

Found this tech article:

http://support.apple.com/kb/TS2757

Mac OS X Server: Server Admin - "There is no server available at the address you entered"
Last Modified: May 26, 2009
Article: TS2757

Symptoms
When trying to connect to a Mac OS X Server via Server Admin, this alert may appear:

"There is no server available at the address you entered"

Products Affected
Mac OS X Server 10.4.7, Mac OS X Server 10.5
Resolution
This may happen for several reasons. Follow the steps below to isolate the issue.

If you're trying to connect via the server's DNS name: Try the IP address instead. If you can connect via the IP address, then correct the DNS entries on your DNS server.
If connecting remotely from another client or server: Make sure TCP port numbers 660, 687, 311 are not being blocked.

Ensure servermgrd is running on the target server. You can use Activity Monitor locally on the server. Or, you can open an ssh session to the server, then execute the following command:

ps aux | grep servermgrd

If servermgrd is not running you can start it by running the following command on the server:
sudo launchctl load -w /System/Library/LaunchDaemons/com.apple.servermgrd.plist
Check Server Admin's preference setting "Require valid digital signature(SSL)". If it's checked (enabled), uncheck it and try again. If it's not enabled, enable it and try again.

This behavior (glacial Server Admin) is usually (but not always) an indication that the DNS environment here is misconfigured or malfunctioning.

Check your DNS server configuration. On the Mac OS X Server server(s) involved (and which may or may not be the DNS server(s) here), check the IP set-up:

+$changeip -checkhostname+

On the client(s) involved with Server Admin, check the translations:

+$ dig -x w.y.z+
+$ dig www.example.com+

Both the forward DNS translations (the "normal" translation stuff) and reverse DNS (rDNS; the DNS translation from the IP address to the host name) both need to work, or (in my experience) Server Admin turns massively sluggish. If you're running your own DNS server(s), then [read this|http://labs.hoffmanlabs.com/node/1436] as a start.

Mac OS X Server Snow Leopard Server 10.6 needs static IP address(es), and needs functional DNS. That can be local (and the default install creates a "runt" DNS for you), or served from another DNS server on your private network or (if you're using public static IP and public DNS) at your ISP DNS server(s).

--- What did NOT work ---

* Ran $changeip -checkhostname and it said there were no changes.
* Nslookup both forward and reverse checkout ok.
* Checked and unchecked SSL within Server Admin prefs and no luck (actually it broke things to turn it on).
* Loaded/Unloaded the plist with no difference.
* Tried serveradmin from the commandline and it just hung.


--- What did work---
* Nothing actually, but I do have a network connection because I can ping, vnc, and mount shares to name a few.

--- What I don't want to do---
I do NOT want to restart the server under any circumstance.

Out of curiosity, how many different folks are in this thread with this question? I see a mix of 10.5 and 10.6 stuff here, and (multiple?) folks asking (related?) questions. I can't tell if this is the same issue (despite the overt similarities) or who has the firewalls dropped and other such, nor whether the access attempt is local or over the network, or otherwise, whether there are multiple NICs present or, well, I'm flying blind. Details would help, and thread-jacking really makes the context ambiguous; do I trust that a previous reply here is reporting identical behavior or not, for instance?

Anyway, for Tim Winningham...

(and if you've not already done so), try dropping the firewall.

And try to use the localhost path if you're using SA on the box, and see if that works.

Some elaboration on the "it broke things" might be useful, too.

Does ping by IP address and by domain name both work?

How many NICs?

Are you using this box as a firewall?

I do NOT want to restart the server under any circumstance.

That's an entirely separate issue that would be worthy of consideration and investigation once this matter is resolved; some degree of redundancy is needed when that "under any circumstance" model arises, whether for cases such as this one or for patches and upgrades and testing, or for the occasional hardware failure.

Edit: I lied. I ended up here linked from another thread. My issue is slightly different - Server Admin is totally unresponsive for me locally and remotely. It's not that certain services don't respond as desired. So, since I guess this does qualify as threadjack, feel free to ignore. ; )

The server in question has dual NICs, but only one is active. Ping to the DNS name and IP of the server works. The server is also serving over AFP and SMB with no interruption. I'm simply unable to resolve the server via Server Admin.

I hope I'm not threadjacking as well here. I appear to have the exact same issue Tim was having, with a similar reticence to reboot the server when everything except Server Admin appears to be working just fine.

Message was edited by: Scarmichael

The same questions I asked Tim Winningham apply here.

Is SA running local? Or remote? If local, to what address are you connecting to SA via? localhost?

How is the second NIC been inactivated?

Where is the DNS server?

What's the DNS configuration here?

It's my experience that there are three potential triggers: bad DNS and particularly bad reverse DNS (which I've seen), and IPv6 weirdnesses and stealth-mode firewall blocks (which I've seen reports).

(I really can't deal with four separate folks asking semi-related questions (or even the same question, if it turns out to be) all at once; that's way more than my brain can track and can manage.)

Well, first off thanks for you help... but you also don't need to feel personally responsible to answer everyone's specific query. It is a community forum after all.

As I stated, I encounter this problem running SA both locally and via a client machine. On the local machine, using the DNS name, IP address and local address all yield the same failure. From a client machine DNS name and IP address yield failure as well.

The second NIC is simply unplugged. I "misspoke" by saying that it was inactive. It's just not being used presently.

The DNS server is on the same network and subnet as the server and client machine. All are on the same LAN.

I'm not sure what specifically you're looking for when you ask for what the DNS configuration is.

As far as I can tell the DNS and reverse DNS check out okay for the machine and firewall is off. However, I've found that trying to load the Sharing Preference panel locks up the System Preferences application, so it may be that the machine is simply having issues. I've scheduled a reboot for the machine tonight, and I imagine that will resolve the issue, at least for now.

I'm always find the "me too" replies useless, but, well, me too. So in the interest of adding data points for comparison sake, in an effort to find commonalities, here goes:

After an indeterminate period of uptime, I can't connect via Server Admin, either on the machine itself (using FQDN, IP, localhost, 127.0.0.1) or remotely. This happens infrequently and seemingly randomly; the servers affected are 10.6.2 Intel Xserves. Typically, they are up for over a month or two, sometimes longer (four months), before this happens.

I've loaded and unloaded the com.apple.servermgrd. I can't telnet to port 311 on the machine itself (no reply) nor can I visit http://fqdn:311 either. The firewall is not active. There are no messages from servermgrd in the system logs (except an expected one generated nightly from a periodic task — and that occurs regardless of the state of Server Admin connections). I don't see anything in CrashReporter logs. I can still use the serveradmin commands via a shell.

I experienced a problem last week on a 10.2.6 box which I thought I would relate. Server Admin suddenly went unresponsive on both of our XServes - from both a client or local attempt. Workgroup Manager worked fine as did Server Monitor. All services were operating normally.

It dawned on me a few days later we'd had a switch failure which necessitated moving the XServes to a different switch. Hadn't thought it would be an issue but seemed to be around the time SA stopped working. Scheduled a restart of one of the servers and lo and behold SA started working again on that machine. Did the same for the second server and everything is now working as expected.

I should point out that prior to restarting we did confirm servermgrd was running on both machines. The only thing I can think of is that SA was running on the servers when the switch failed which screwed something. We did get the normal "An error has occurred" message from SA when the switch failed but we've seen that message fairly regularly over the years and it's never affected the ability of SA to simply reconnect.

FWIW, if you're working with managed switches, the first stop is the switch security and MAC settings; the port security and related. It's common to see issues (eg: security lock-outs, vLAN configuration confusion, etc) with some managed switches as a result of the presence of the LOM on the connection.

MrHoffman wrote:
FWIW, if you're working with managed switches, the first stop is the switch security and MAC settings; the port security and related. It's common to see issues (eg: security lock-outs, vLAN configuration confusion, etc) with some managed switches as a result of the presence of the LOM on the connection.

Interesting, thanks. We only have LOM on one machine but if you're suggesting that could affect the switch then it's feasible it could stop both machines connecting. Having said that, I still find it strange the local connection wouldn't work, and WGM continued happily - both remotely and locally.

Hey there everyone!

I am having the same issues as everyone here. I cannot connect with Server Admin to my 10.6.2 Server. Some Stats:
* Server has both NIC's active, one private one public.
* All service are running just fine, SSH, ARD, WEB, etc.
* I can ping both internal and external NIC's.
* No configuration changes
* servermgrd is running as root

Doing a netstat doesn't show any port 311 listening or open. So this tells me that the process isn't even listening. In looking at the logs for servermgrd I noticed that back in Feb there are a series of:

#Start-Date: 2010-02-20 08:27:27 MST
#Fields: date time s-comment
2010-02-20 08:27:27 MST Initalizing plugins
2010-02-20 08:27:31 MST Listening for requests

Well when I started not being able to connect to the server with SA, was about Feb 24. On the log on that day, and all entries after that it shows:

#Start-Date: 2010-02-25 10:56:34 MST
#Fields: date time s-comment
2010-02-25 10:56:34 MST Initalizing plugins

NO listening for requests anymore. So somehow this sucker doesn't want to listen anymore. I have had this problem in the past, and a server reboot was all ways a good cure. I'm pretty sure I could reboot her now, but I'd really rather find a solution to the problem.

Does this shed some light on it for anyone? Any ideas out there?

Thanks for the community,
Eric

I actually created another post for this, but I found this thread after searching for a couple hours. I am having the same issue, however, a server restart doesn't fix it for me. And Apple's article fix doesn't help either.

Here are my logs:
After a restart it appears servermanager can't connect via port 311:
{quote}
4/8/10 6:18:54 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:18:54 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:18:54 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:06 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:06 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:06 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:18 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:18 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:18 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:30 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:30 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:30 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(interactive,10.1.184.26,311): 61
4/8/10 6:19:40 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(localhost,127.0.0.1,311): 61
4/8/10 6:19:40 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(localhost,127.0.0.1,311): 61
4/8/10 6:19:40 PM /Applications/Server/Server Monitor.app/Contents/MacOS/Server Monitor[978] [978] ServerManager session failed in connect(localhost,127.0.0.1,311): 61
{quote}

And here are my servermgrd logs:
{quote}
4/6/10 12:14:12 PM servermgrd[65184] Allocated size has grown to 4M. Number of allocations is -5890. Exiting to clear possible memory leak.
4/6/10 12:14:12 PM com.apple.launchd[1] (com.apple.servermgrd[65184]) Exited with exit code: 12
4/6/10 12:14:13 PM servermgrd[98795] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/6/10 12:14:13 PM servermgrd[98795] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/7/10 12:14:13 AM servermgrd[98795] Allocated size has grown to 4M. Number of allocations is -10672. Exiting to clear possible memory leak.
4/7/10 12:14:13 AM com.apple.launchd[1] (com.apple.servermgrd[98795]) Exited with exit code: 12
4/7/10 12:14:14 AM servermgrd[43681] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/7/10 12:14:14 AM servermgrd[43681] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/7/10 8:16:47 AM servermgrd[43681] [43681] error in getAndLockContext: flock(servermgr_ftp) FATAL time out
4/7/10 8:16:47 AM servermgrd[43681] [43681] process will force-quit to avoid deadlock
4/7/10 8:16:47 AM com.apple.launchd[1] (com.apple.servermgrd[43681]) Exited with exit code: 1
4/7/10 12:59:11 PM com.apple.launchd[1] (com.apple.servermgrd[66064]) Exited: Killed
4/8/10 6:01:34 PM servermgrd[57] servermgr_web: Created WebDAV lock directory /var/run/davlocks
4/8/10 6:01:35 PM servermgrd[57] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/8/10 6:01:35 PM servermgrd[57] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/8/10 6:01:36 PM servermgrd[57] servermgr_info: mark running_servicesconfigured(): marked running services configured: com.apple.ServerAdmin.Web, com.apple.ServerAdmin.AppleFile, com.apple.ServerAdmin.MySQL, com.apple.ServerAdmin.FTP, com.apple.ServerAdmin.Windows
4/8/10 6:04:11 PM servermgrd[57] servermgr_jabber[N]: Network change ignored - no change in host(s) or address(s).
4/8/10 6:06:21 PM servermgrd[57] servermgr_jabber[N]: Network change ignored - no change in host(s) or address(s).
4/8/10 6:07:44 PM servermgrd[57] servermgr_jabber[N]: Network change ignored - no change in host(s) or address(s).
4/8/10 6:07:57 PM servermgrd[57] servermgr_jabber[N]: Network change ignored - no change in host(s) or address(s).
4/8/10 6:15:58 PM servermgrd[61] servermgr_web: Created WebDAV lock directory /var/run/davlocks
4/8/10 6:15:59 PM servermgrd[61] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/8/10 6:15:59 PM servermgrd[61] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/8/10 6:24:17 PM servermgrd[61] servermgr_jabber[N]: Network change ignored - no change in host(s) or address(s).
4/9/10 6:15:59 AM servermgrd[61] servermgr_info: unexpected Software Update state: crashed
4/9/10 9:15:59 AM servermgrd[61] Allocated size has grown to 4M. Number of allocations is -8113. Exiting to clear possible memory leak.
4/9/10 9:15:59 AM com.apple.launchd[1] (com.apple.servermgrd[61]) Exited with exit code: 12
4/9/10 9:16:00 AM servermgrd[14887] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/9/10 9:16:00 AM servermgrd[14887] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/9/10 12:29:12 PM com.apple.launchd[1] (com.apple.servermgrd[14887]) Exited: Killed
4/9/10 12:29:12 PM servermgrd[18682] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/9/10 12:29:12 PM servermgrd[18682] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/9/10 2:22:21 PM sudo[21225] admin : TTY=ttys000 ; PWD=/Users/admin ; USER=root ; COMMAND=/System/Library/CoreServices/ServerManagerDaemon.bundle/Contents/MacOS/ servermgrd servermgrd -x
4/9/10 2:22:22 PM com.apple.ReportCrash.Root[21231] 2010-04-09 14:22:22.324 ReportCrash[21231:2803] Saved crash report for servermgrd[21230] version ??? (???) to /Library/Logs/DiagnosticReports/servermgrd 2010-04-09-142222localhost.crash
4/9/10 2:23:09 PM com.apple.launchd[1] (com.apple.servermgrd[18682]) Exited: Terminated
4/9/10 2:23:10 PM servermgrd[21234] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/9/10 2:23:10 PM servermgrd[21234] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/9/10 2:23:35 PM com.apple.launchd[1] (com.apple.servermgrd[21234]) Exited: Terminated
4/9/10 2:23:35 PM servermgrd[21260] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/9/10 2:23:35 PM servermgrd[21260] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
4/9/10 2:28:56 PM sudo[21990] admin : TTY=ttys000 ; PWD=/Users/admin ; USER=root ; COMMAND=/bin/launchctl unload -w /System/Library/LaunchDaemons/com.apple.servermgrd.plist
4/9/10 2:29:16 PM sudo[21991] admin : TTY=ttys000 ; PWD=/Users/admin ; USER=root ; COMMAND=/bin/launchctl load -w /System/Library/LaunchDaemons/com.apple.servermgrd.plist
4/9/10 2:29:16 PM servermgrd[21992] servermgr_ipfilter:ipfw config:Notice:Flushed IPv4 rules
4/9/10 2:29:17 PM servermgrd[21992] servermgr_ipfilter:ipfw config:Notice:Flushed IPv6 rules
{quote}

Can't I just activate port 311? Why is this happening? Why would the port just suddenly not be active anymore?

Thanks!

The LOM is not particularly related to WGM; the LOM is active and visible and operates at a lower level.

Many (most) managed switches operate based detecting and processing the Media Access Control (MAC, and not "Mac") address of the connected client in some fashion. Managed switches can be "confused" when they detect a MAC address change (and can interpret that as a cable swap), and managed switches can optionally also be configured to lock out unrecognized or unexpected MAC addresses.

Depending on the configuration, the Mac has a MAC, and the LOM adds a second MAC. On the same network connection. And two MACs on one wire can sometimes "befuddle" a managed switch.

Depending on the particular switch and your local networking configuration and requirements, it can sometimes be expeditious to connect the Mac via one NIC and the LOM via the other NIC, too.

Unmanaged switches generally don't "care" about these details.

Please see my post here for my fix:
http://discussions.apple.com/thread.jspa?threadID=2393971&tstart=0&start=30

Hopefully it will work for you as well!!!