| |
|
Alert Type: | New |
Bug Id: | CSCuy86016 | Title: | SMD proc may be held-down with EAP-FAST local & Multi Auth Host Mode |
|
Status: | Open |
|
Severity: | 1 Catastrophic |
Description: | Symptom:
The process smd has been held down with EAP-FAST local authentication and Multi-Auth Host Mode. It will happen when concerned interface have more than one session.
Conditions:
This issue will be seen only with local authentication (EAP-FAST) & multi-auth host mode only.
Workaround:
There is no workaround.
|
|
Last Modified: | 28-APR-2016 |
|
Known Affected Releases: | Denali-16.3.1 |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuw01266 | Title: | ffm crash when adding 3.3.x switch to 3.6.3 stack |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
switch with 3.3.x release crashes when added to the stack running 3.6.3 release
Conditions:
Member switch running 3.3.x crashes with Feature Forwarding Manager service upon joining the stack
Workaround:
Install 3.6.3 on a switch before adding the switch to the stack
Further Problem Description:
This behavior is not seen when switch with 3.6.3 is added to the stack already running 3.6.3, switches loaded with same version will never hit this issue.
|
|
Last Modified: | 07-APR-2016 |
|
Known Affected Releases: | 3.6 |
|
Known Fixed Releases: * | 15.2(2)E4, 15.2(3)E3, 3.6(4)E, 3.7(3)E |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuz29721 | Title: | Memory leak under HTTP EPM Redirect Daemon |
|
Status: | Open |
|
Severity: | 2 Severe |
Description: | Symptom:
Memory leak under malloc lite if turned on. If not, the memory will be held under *Dead*
Conditions:
HTTP redirection on.
Workaround:
none at this time.
Further Problem Description:
|
|
Last Modified: | 26-APR-2016 |
|
Known Affected Releases: | 3.7(3)E |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy81074 | Title: | Stack Manager requested reload due to ASIC ballot failure |
|
Status: | Open |
|
Severity: * | 2 Severe |
Description: | Symptom:This issue has been seen in a 3850 STACK running 3.6.4E. The switch will boot after upgrading to 3.6.4E (in the test case from 3.6.3E). and it will work well in a random interval of 2 to 10 minutes. After this, the switch will display the "Stack Manager requested reload due to ASIC ballot failure" message and will reload. This behavior will keep repeating itself.
Conditions:3850 Stack running 3.6.4E.
Workaround:When the switches are standalone, that is NO stack, the issue is not presented and the switches run steadily.
More Info:Migrating from 3.6.3E to 3.6.4E , although we believe that the previous release does not set precedence. This issue was NOT seen in 3.6.3E in the same stack.
This issue occurred in one out of 6 stacks in our lab.
|
|
Last Modified: | 06-APR-2016 |
|
Known Affected Releases: | 15.2(2)E4 |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuy16820 | Title: | 'show SUDI' cmd: Unsupported for c3850/c3650 switches |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Secure Unique Device Identifier Sudi Support is expected for NG3K Latest Certificate. "show crypto pki certificate" will show the latest SHA1 certificate always as cisco SUDI legacy id, and SHA2 as cisco sudi id. Expecting to see the latest certificate as cisco sudi id without legacy
Conditions:
While checking board with only sha1 certificate, it will show cisco sudi legacy
while checking board with both sha1 and sha2 certificate, sha1 will show as legacy and sha2 as cisco sudi id
Workaround:
None
Further Problem Description:
Issue fixed in Denali 16-3, Customers can upgrade to that version when available for this fix.
|
|
Last Modified: | 06-APR-2016 |
|
Known Affected Releases: | 1.0 |
|
Known Fixed Releases: | 16.3(0.119) |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy85838 | Title: | Clients stuck 'Method status list: empty' after SSO->shut/no shut helps |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Some of the clients are stuck in Unauth forever with 'Method status list: empty'.
Conditions:
The issue happens intermittently after reload of the stack or switchovers.
Workaround:
shut and no shut of the interface.
Further Problem Description:
|
|
Last Modified: | 21-APR-2016 |
|
Known Affected Releases: | 16.2.1 |
|
Known Fixed Releases: * | 15.2(2)E2, 15.2(2)E3, 15.2(2)E4, 15.2(3)E, 15.2(3)E1, 15.2(3)E2, 15.2(3)E3, 15.2(4)E, 15.2(4)E1, 15.5(1)S |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuj10443 | Title: | Standby sw crash@crypto_engine/sw/src/keylib/lib_key_storage.c:646 |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Switch crashes when being added to stack after switchoer
Conditions:
This occurs in switchover scenarios
Workaround:
Reload the complete stack and boot again.
Further Problem Description:
|
|
Last Modified: | 21-APR-2016 |
|
Known Affected Releases: | 15.0(11.98)EMP, 15.0(12.3)EMP, 15.2(2.2.94)E |
|
Known Fixed Releases: * | 12.2(60)EZ2, 12.2(60)EZ3, 15.0(1)EZ, 15.0(1)EZ1, 15.0(12.14)EZD, 15.0(14.1)TSR, 15.0(14.49)EZD, 15.0(2)EX5, 15.0(2)EX7, 15.0(2)SE7 |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy64839 | Title: | SMD Crash on Active Switch during bootup |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
SMD Crash observed
Conditions:
Probable Conditions are when APs flap at join
Workaround:
Reload the Stack.
Further Problem Description:
|
|
Last Modified: | 21-APR-2016 |
|
Known Affected Releases: | 16.2.1, Denali-16.2.1 |
|
Known Fixed Releases: * | 15.2(1)E, 15.2(1)E1, 15.2(1)E2, 15.2(1)E3, 15.2(1)EY, 16.2(0.363), 16.3(0.147) |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy73209 | Title: | Client association failed in native-profiling test |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Symptom:
Wireless client is not able to join to a WLAN which has native profile and AAA override enabled
Conditions:
Conditions:
WLAN with native profile policy and client is able to join.
Enable AAA override on the same WLAN; client is not able to join
Workaround:
Workaround:
Run "clear access-session cache" command.
Further Problem Description:
Further Problem Description:
|
|
Last Modified: | 21-APR-2016 |
|
Known Affected Releases: | 16.2(0) |
|
Known Fixed Releases: * | 16.2(0.363), 16.3(0.145) |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCux24057 | Title: | Install/upgrade on a specific non-active switch will fail |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
When doing a superpackage install on a specific member, other than active, using the "request platform software package install" command will result in an error message and install will fail.
Conditions:
This only occurs when doing superpackage install on a specific non-active switch.
Issue is not seen when using "switch all", or when the switch specified is the current active.
Workaround:
The expansion of the superpackage to member will be successful, only the install portion will have failed. In such cases, user can reissue the install command on that specific switch but instead of specifying the superpackage as the filename, they can use *.pkg expression to denote the necessary subpackages instead.
Example:
Superpackage cat3k_caa-universalk9.BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA.bin
expands to the following subpackages:
cat3k_caa-rpbase.BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA.pkg
cat3k_caa-srdriver.BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA.pkg
cat3k_caa-wcm.BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA.pkg
cat3k_caa-webui.BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA.pkg
Build string in this case would be BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA
So user can run install on all the extracted subpackages using:
request platform software package install switch file flash:*BLD_V161_0_THROTTLE_LATEST_20151116_230450_2.SSA.pkg
Further Problem Description:
|
|
Last Modified: | 10-APR-2016 |
|
Known Affected Releases: | 16.3(0) |
|
Known Fixed Releases: * | 16.1(2.43), 16.2(0.350), 16.3(0.140) |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuy70285 | Title: | Perennial traceback after SSO and some windows client not able to rejoin |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
After certain number of Switchcovers, there could be a chance that perennial traceback being observed. Some clients may not be able to join back.
Conditions:
Certain number of Switchovers, perennial traceback could be seen.
Workaround:
No workaround
Further Problem Description:
|
|
Last Modified: | 06-APR-2016 |
|
Known Affected Releases: | Denali-16.2.1 |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy51956 | Title: | AAA server moves to DEAD state even though CTS server is ALIVE |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Dot1x accounting fails when trying to send to CTS enabled radius server due to invalid authenticator, because of this AAA server might go to DEAD state intermittently.
Conditions:
When we have AAA accounting enabled along with CTS authorization , might run into this issue
Workaround:
Remove the AAA related Accounting configuration to over come this issue.
Further Problem Description:
|
|
Last Modified: | 14-APR-2016 |
|
Known Affected Releases: | Denali-16.2.1 |
|
Known Fixed Releases: * | 16.2(0.355), 16.3(0.113) |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuu18029 | Title: | 3650/3850 May Experience RP Protocol Flaps with Aggressive Timers. |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: * | Symptom:
The 3650 and 3850 switches may experience Routing Protocol Flaps when using aggressive timers.
Conditions:
The problem has been seen on the NGWC switches with RP protocols using 3 seconds or less for dead/hold-down timers.
Workaround:
It is recommended to use the default routing protocol timers. If faster convergence is required do not use anything less than 10 second dead/hold-down to avoid this defect.
Further Problem Description:
|
|
Last Modified: | 18-APR-2016 |
|
Known Affected Releases: | 15.2(2.0)E |
|
Known Fixed Releases: | 15.2(2)E4, 15.2(3)E3, 3.6(4)E, 3.7(3)E |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuw55669 | Title: | Crash is seen in iosd on switch and auth-mgr |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Switch with IOSD crashes in auth-manager with this symptom:
IOSD-EXT-SIGNAL: Segmentation fault(11), Process = EPM MAIN PROCESS
Conditions:
The system crashed due to stack overflow caused by loop calls between PI sanet and vlan group plugin on switch platform while dumping sanet trace messages in the case when handling dynamic vlan updates for a given vlan on standby switch.
Workaround:
None
Further Problem Description:
|
|
Last Modified: | 02-APR-2016 |
|
Known Affected Releases: | 3.6(0) |
|
Known Fixed Releases: * | 15.2(2)E4, 15.2(3)E3, 15.2(4)E1, 15.2(4)EA1, 15.2(4.0.95)EA1, 15.2(4.0.95a)E, 15.2(4.1.25)EA1, 15.2(4.1.5a)E, 15.2(4.2.1)EA2, 15.2(5.0.38)E |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCux22760 | Title: | On reload, 1G SFP is connected/up on 3750x while 3850 is notconnect/down |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
Topology: 3850-12S downlink port connected to uplink port of 3750X using 1G SFP.
Now on reload of 3850, 1G SFP is connected/up on 3750x while 3850 is notconnect/down.
Conditions:
Topology: 3850-12S downlink port connected to uplink port of 3750X using 1G SFP.
Now on reload of 3850, 1G SFP is connected/up on 3750x while 3850 is notconnect/down.
Workaround:
No Workaround
Further Problem Description:
The defect is fixed in Polaris dev and amur mr4 (3.6.4) and beni mr4 (3.7.4)
|
|
Last Modified: | 27-APR-2016 |
|
Known Affected Releases: | 15.2(2)E3 |
|
Known Fixed Releases: * | 15.2(2)E4, 16.1(2.62), 16.2(0.199), 3.6(4)E |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCux77511 | Title: | 3850: Webauth not working when incorrect username/password entered |
|
Status: | Fixed |
|
Severity: | 2 Severe |
Description: | Symptom:
For a switch configured to run with WEBAUTH and MAB, when a user enters invalid username/password on logon page, the failed window opens. However, nothing is displayed and it just sits there trying to connect to the page you want to get to and doesn't connect to anything. After this, the connection is unresponsive and the login screen doesn't function. The connection becomes responsive again after resetting the connection (unplugging/plugging the Ethernet cable) but WEBAUTH only works when a valid username/password is entered.
Conditions:
Workaround:
Further Problem Description:
|
|
Last Modified: | 20-APR-2016 |
|
Known Affected Releases: | 15.2(3)E |
|
Known Fixed Releases: * | 15.2(2)E4, 15.2(5.1.8)E, 15.2(5.5.66)E |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuz29932 | Title: | cat3k when connected to 8510 WLC fails to forward capwap traffic |
|
Status: | Open |
|
Severity: | 2 Severe |
Description: | Symptom:
C3850 fails to forward CAPWAP traffic to 8510 WLC.
Conditions:
The issue is seen when C3850 is directly connected to 8510 WLC.
Workaround:
The workaround is to ping WLC from C3850.
Further Problem Description:
|
|
Last Modified: | 26-APR-2016 |
|
Known Affected Releases: | Denali-16.3.1 |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCux24515 | Title: | "%Failed to update configuration" message seen on boot-up sometimes |
|
Status: | Fixed |
|
Severity: | 3 Moderate |
Description: | Symptom:
After booting up, the error "%Failed to update configuration in the database. Please revert" is printed to console.
Conditions:
In rare condition when "attribute password PASSWORD" contains the characters 3F, this error is printed. This is however, non functional impacting
Workaround:
None
Further Problem Description:
|
|
Last Modified: | 27-APR-2016 |
|
Known Affected Releases: * | Denali-16.1.1 |
|
Known Fixed Releases: | 16.1(1.50), 16.1.2, 16.2(0.201), Denali-16.1.2 |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCum71431 | Title: | Cat3850 FED Reload when trying to being up switch stack |
|
Status: | Terminated |
|
Severity: | 3 Moderate |
Description: | Symptom:
In an environment where downloadable ACLs are used a Cat3850 may report a FED crash.
Conditions:
This crash is triggered due to ACL names longer than 64 characters. Internally, several strings are added to the ACL name, and if the aggregate exceeds 63 characters this crash would be seen.
As an example, consider an ACL configured in an ACS server as "this-acl-is-22-letters".
- ACS adds characters to the beginning and end of this ACL (which can be seen in "show auth sess int detail
- the switch adds characters to the beginning and end of this ACL
#sh auth sess int gi 2/11 details
Interface: GigabitEthernet2/11
...
Authorized By: Authentication Server
ACS ACL: xACSACLx-IP-this-acl-is-22-letters-520492d0
...
Any port where a device receives an ACL which is longer then 64 characters will be truncated and this same truncated string will share the ACL programming with other ports in same state and trigger the issue.
Workaround:
Shorten ACL names to less than 64 characters. As additional character are padded towards the beginning/end of the ACL name those characters should be factored in when renaming the ACLs.
Further Problem Description:
This issue is fixed in 3.6.0E / 15.2(2)E and later releases.
|
|
Last Modified: | 26-APR-2016 |
|
Known Affected Releases: | 15.0(1)EZ |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy54762 | Title: | IPv4 DHCP traffic is intercepted by L3 SVI interface with ACL log |
|
Status: | Fixed |
|
Severity: | 3 Moderate |
Description: * | Symptom:
When the IPv4 unicast traffic should be switched by 3850, it is intercepted and copied to L3 SVI interface on IN direction. If there is an ACL on IN direction applied to SVI interface, the log message is printed. However, traffic still is L2 switched to the destination, while the copy of the traffic addressed to L3 SVI is dropped.
Example:
*Feb 29 00:06:26.087: %SEC-6-IPACCESSLOGP: list INNER denied udp 10.1.1.1(67) (Vlan500 5ca4.8afb.6dd0) -> 10.1.1.8(68), 6 packets
*Feb 29 00:11:26.088: %SEC-6-IPACCESSLOGP: list INNER denied udp 10.1.1.1(67) (Vlan500 5ca4.8afb.6dd0) -> 10.1.1.8(68), 6 packets
Both 10.1.1.1(relay agent) and 10.1.1.8(client) do not belong to the L3 SVI interface, so the traffic should be switched using L2 mac-address-table only and not printing Log messages
Conditions:
Only DHCP unicast conversations are affected, other traffic is not affected
Workaround:
None
Further Problem Description:
|
|
Last Modified: | 22-APR-2016 |
|
Known Affected Releases: | 15.2(3.7.2) |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy79579 | Title: | SMD leak while testing CWA disconnect&reauth |
|
Status: | Fixed |
|
Severity: | 3 Moderate |
Description: | Symptom:
SMD memory leak after central webauth clients associated and disconnected
Conditions:
central webauth clients
Workaround:
NA
Further Problem Description:
|
|
Last Modified: | 22-APR-2016 |
|
Known Affected Releases: | Denali-16.2.1 |
|
Known Fixed Releases: * | 16.2(0.364), 16.3(0.145) |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCul67169 | Title: | Fed Crash on NG3K when trying to authenticate a supplicant using 802.1x |
|
Status: | Fixed |
|
Severity: | 3 Moderate |
Description: | Symptom:
In an environment where downloadable ACLs are used a Cat3850 may report a FED crash.
Conditions:
This crash is triggered due to ACL names longer than 64 characters. Internally, several strings are added to the ACL name, and if the aggregate exceeds 63 characters this crash would be seen.
As an example, consider an ACL configured in an ACS server as "this-acl-is-22-letters".
- ACS adds characters to the beginning and end of this ACL (which can be seen in "show auth sess int detail
- the switch adds characters to the beginning and end of this ACL
#sh auth sess int gi 2/11 details
Interface: GigabitEthernet2/11
...
Authorized By: Authentication Server
ACS ACL: xACSACLx-IP-this-acl-is-22-letters-520492d0
...
Any port where a device receives an ACL which is longer then 64 characters will be truncated and this same truncated string will share the ACL programming with other ports in same state and trigger the issue.
Workaround:
Shorten ACL names to less than 64 characters. As additional character are padded towards the beginning/end of the ACL name those characters should be factored in when renaming the ACLs.
Further Problem Description:
|
|
Last Modified: | 18-APR-2016 |
|
Known Affected Releases: | 15.0(1)EZ |
|
Known Fixed Releases: | 15.2(2)E |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuy24630 | Title: | Not able to open webauth page in client with ipv6 address. |
|
Status: | Fixed |
|
Severity: | 3 Moderate |
Description: | Symptom:
Webauth redirection not happening for IPv6 only wireless clients to do web authentication
Conditions:
3850 configured with IPv6 virtual-ip address in web-auth parameter-map.
1. Wireless client with only IPv6 address
2. Open IPv6 web page or domain in browser
Workaround:
NA
Further Problem Description:
|
|
Last Modified: | 14-APR-2016 |
|
Known Affected Releases: | Denali-16.2.1 |
|
Known Fixed Releases: * | 16.2(0.355), 16.3(0.156) |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuy34177 | Title: | Need 5508 to support sleeping client as single Anchor with NGWC |
|
Status: | Open |
|
Severity: | 3 Moderate |
Description: | Symptom:
Sleeping Clients will not be seen in sleeping Client list on Guest Anchor (5508). When the clients wake up, they will have to re-authenticate.
Conditions:
When Clients are joined in Foreign Anchor (NGWC) <--> Guest Anchor (5508) Scenario.
Workaround:
No Workarounds for now.
Further Problem Description:
|
|
Last Modified: | 06-APR-2016 |
|
Known Affected Releases: | 16.2.0 |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuy79465 | Title: | Prime - Unable to modify Transmitted Power Threshold for TPC |
|
Status: | Open |
|
Severity: | 3 Moderate |
Description: | Symptom:
Unable to make configuration changes on Prime to the Transmit Power Threshold for TPC.
Conditions:
Attempting to modify the Transmit Power Threshold via Prime.
Workaround:
Make changes on Polaris controller web interface (webui).
Further Problem Description:
|
|
Last Modified: | 06-APR-2016 |
|
Known Affected Releases: | Denali-16.2.1 |
|
Known Fixed Releases: | |
|
|
| |
| |
|
Alert Type: | Updated * |
Bug Id: | CSCuv03683 | Title: | %Failed to determine package type when sw install from non-Active flash |
|
Status: | Fixed |
|
Severity: | 3 Moderate |
Description: | Symptom:
When a user attempts to do a software install where you specify the source image from a non-Active flash switch, it will fail with a "% Failed to determine package type for flash-X:image.bin"
Conditions:
Customer has to be trying to install an image from a non-Active flash in order to see the issue.
Workaround:
The workaround would be to copy the image to the Active flash and then issue the software install from that flash instead of a non-Active flash - software install file flash:image.bin.
Further Problem Description:
|
|
Last Modified: | 05-APR-2016 |
|
Known Affected Releases: | 15.2(3.7.1E) |
|
Known Fixed Releases: * | 15.2(2)E4, 15.2(3)E3, 3.6(4)E, 3.7(3)E |
|
|
| |
| |
|
Alert Type: | New |
Bug Id: | CSCuz47111 | Title: | Disabling Option 66&67Stack 3850Suddenly a member reload config mismatch |
|
Status: | Open |
|
Severity: | 3 Moderate |
Description: | Symptom:
Suddenly the standby switch reload after disable commands "option 66" and "option 67" and write the config.
Conditions:
w-3850-2#conf t
Enter configuration commands, one per line. End with CNTL/Z.
w-3850-2(config)#ip dhcp pool Vlan_16
w-3850-2(dhcp-config)#no option 66
w-3850-2(dhcp-config)#no option 67
w-3850-2(dhcp-config)#
w-3850-2(dhcp-config)
|
没有评论:
发表评论