When executing an Overlay, the error message "can't connect; disconnect" is seen


<< Back to Knowledge Search

Solution

Overview

When executing an overlay you see this message error message:  "can't connect; disconnect" is reported and Overlay fails.  This occurs even though you see a 'green light' on the Director UI for this device. When Director tries to push the configuration to the SG using the overlay file, sometimes the configuration push will not be completed.

Connections to the SG using ssh will sometimes hang, especialy when attempting to push a configuration file to it.

This occurs when using Director SGME 5.5.1.1/5.5.2.3 against the to ProxySG SGOS 5.4.5.1   and some early SGOS 6.1.X versions.

Cause
Resolution

This is because the SSH session between director and SG will be hung(inactive). There was an incompatibilty between Director SGME 5.5.1.1 / 5.52.3 and ProxySG 5.4.5.1, and some early Proxy versions 6.2.X releases.

FIX: To avoid this issue, we recomend you upgrading your ProxySG to 5.4.6.1.  and/or SGOS 6.1.5.1 GA released on 8/8/2011

An example of what you see on the Directors /var/log/messages screen is below.

Feb  2 12:04:13 director ccd: <ccd.notice> Device wilpxy02: disconnected, Reason Not Connected - Device Disconnected
Feb  2 12:04:13 director dmd: <dmd.warn> Device ID "wilpxy02" error: Not Connected - Device Disconnected
Feb  2 12:04:13 director dmd: <dmd.notice> Health state for metric"wilpxy02/12" "ok" changed to "disconnected", reason: "Device connection"
Feb  2 12:04:13 director configd: <configd.notice> CCD lost connection to device "wilpxy02": (5007103240) Not Connected - Device Disconnected 
Feb  2 12:04:13 director dmd: <dmd.crit> Insert of Alert record is failed
Feb  2 12:04:13 director dmd: <dmd.crit> PQResultErrorMessage::ERROR:  relation "alertrecord" does not exist 

For SGs on 6.1x symptoms often include a timeout on the Director/SG appliance connection, and manifest "timeout"  as per the below example of the the /var/log/messages file.

Jul 11 09:30:00 vmdirector schedulerd: <schedulerd.notice> Executing Job "sync"execution 1310376600

Jul 11 09:30:00 vmdirector runner[6441]: <runner.notice> Processing command: 1310376600501105:remote-config overlay "policy" refresh Jul
11 09:30:00 vmdirector configd: <configd.notice> Importing overlay <policy> from HPSN2-WCF-PO63TH-PX1

Jul 11 09:30:09 vmdirector runner[4933]: <runner.crit> Timeout expired while waiting for message Jul 11 09:30:09
vmdirector runner[4933]: <runner.crit> sd_rp_configd_async_read(),
sd_runner.c:846, build 000000: Error 1 returned, bailing out.
Jul 11 09:30:09 vmdirector runner[4933]: <runner.crit> cli_send_configd_msg(),
libcli_main.c:457, build 000000: Error 1 returned, bailing out.
Jul 11 09:30:09 vmdirector runner[4933]: <runner.crit> Error 1 returned, bailing out.
Jul 11 09:30:09 vmdirector runner[4933]: <runner.crit>
libcli_process_command_wrapper(), libcli_main.c:210, build 000000: Error 4 returned, bailing out.

 

NOTE1:  For information on how to access the Directors /var/log/messages file, see KB4178

NOTE2: For information on what others errors we see in a profile push, see 000016302

NOTE3: For information on what CLI commands we use to push a profile, see 000014751

Workaround
Additional Information
Bug Number
InQuira Doc IdKB4315
Attachment

Article Feedback

Hide Properties
First Published      10/01/2014
Last Modified      10/01/2014
Last Published      10/01/2014
Article Audience
Product      Director 510
Software      SGME 5.5, SGOS 5.4
Article Number      000015709
Summary     
Was this helpful?
Comments:
 
Previous MonthNext Month
SunMonTueWedThuFriSat