Menu
▾
▴
Re: [linux-iscsi-users] Connection to Discovery Address failed [EMCAX150SCi on RHEL4]
|
From: Matthew H. <ma...@da...> - 2007-01-03 16:27:01
|
No, not yet. EMC is still looking into it the last I talked with them =
before the holidays.
I also can get it to work as you have as well. However, it seems to =
start acting flakey (large 30-60+ second timeouts for disk activity) =
under moderate to heavy I/O usage with the Linux iSCSI connections. I =
don't have any physical documentation of this occurring other than =
watching it stall (as it doesn't seem to be writing it to the logs) or =
time to look into it further at the moment.
I haven't experienced these issues with the Microsoft iSCSI initiators =
under very similar usage patterns. At this point I've stopped using it =
except for light use (and continued testing) on Linux until I feel more =
confident in it working as I would expect.
I apologize that I can't be of much help in this matter at the moment.
-----Original Message-----
From: lin...@li... =
[mailto:lin...@li...] On Behalf Of =
S=F8ren Hjarlvig
Sent: Wednesday, January 03, 2007 3:51 AM
To: lin...@li...
Subject: Re: [linux-iscsi-users] Connection to Discovery Address failed =
[EMCAX150SCi on RHEL4]
> I'm getting this error in my event logs almost precisely every 30
> seconds. The iscsi connection fails to the discovery address, and then
> is immediately made again.
>=20
> Dec 14 15:48:52 chlorine iscsid[19813]: Connection to Discovery =
Address
> 192.168.15.2 failed=20
> Dec 14 15:48:52 chlorine iscsid[19815]: Connection to Discovery =
Address
> 192.168.15.3 failed=20
> Dec 14 15:48:52 chlorine iscsid[19813]: Connected to Discovery Address
> 192.168.15.2=20
> Dec 14 15:48:52 chlorine iscsid[19815]: Connected to Discovery Address
> 192.168.15.3
>=20
> Does anyone have any ideas what I can do to remedy this? All RHEL4U4
> hosts that connect to this target exhibit this behavior, while the
> windows hosts do not.
Hello Matthew!
Just wondering if you have found a solution ?
I'm having the exact same problem on a debian sarge box (kernel
2.6.8-3-686-smp, SFNet iSCSI Driver Version ... 4:0.1.10, Dell/EMC
AX150SCi):
Jan 3 00:22:14 rocksteady iscsid[2408]: Connection to Discovery Address
10.10.1.10 failed
Jan 3 00:22:14 rocksteady iscsid[2408]: Connected to Discovery Address
10.10.1.10
Jan 3 00:22:45 rocksteady iscsid[2408]: Connection to Discovery Address
10.10.1.10 failed
Jan 3 00:22:45 rocksteady iscsid[2408]: Connected to Discovery Address
10.10.1.10
I have tried the Continuous=3Dno/yes option and setting the global idle
timeout to something below 30 (20):
iscsi-ls -c:
*************************************************************************=
******
SFNet iSCSI Driver Version ... 4:0.1.10 ( 8-Oct-2004)
*************************************************************************=
******
TARGET NAME : iqn.1992-04.com.emc:ax.fcnpr062301759.a0
TARGET ALIAS :
HOST NO : 3
TARGET ADDRESS : 10.10.1.10:3260
SESSION STATUS : CONNECTION ON 10.10.1.10:3260
ESTABLISHED AT Tue Jan 2 23:20:45 2007
NO. OF PORTALS : 1
PORTAL ADDRESS 1 : 10.10.1.10:3260,1
SESSION ID : ISID 00023d000001 TSIH a197
TARGET CONFIGURATION KEYS :
-------------------------
INITIAL R2T : Yes
IMMEDIATE DATA : No
HEADERDIGEST : NONE
DATA DIGEST : NONE
FIRSTBURSTLENGTH : 0
MAXBURSTLENGTH : 262144
MAXRECVDATASEGMENTLENGTH: initiator - 131072
target - 65536
LOGIN TIMEOUT : 15
AUTH TIMEOUT : 45
ACTIVE TIMEOUT : 5
IDLE TIMEOUT : 60
PING TIMEOUT : 5
*************************************************************************=
******
TARGET NAME : iqn.1992-04.com.emc:ax.fcnpr062301759.a1
TARGET ALIAS :
HOST NO : 4
TARGET ADDRESS : 10.10.2.10:3260
SESSION STATUS : CONNECTION ON 10.10.2.10:3260
ESTABLISHED AT Tue Jan 2 23:20:45 2007
NO. OF PORTALS : 1
PORTAL ADDRESS 1 : 10.10.2.10:3260,2
SESSION ID : ISID 00023d000001 TSIH a198
TARGET CONFIGURATION KEYS :
-------------------------
INITIAL R2T : Yes
IMMEDIATE DATA : No
HEADERDIGEST : NONE
DATA DIGEST : NONE
FIRSTBURSTLENGTH : 0
MAXBURSTLENGTH : 262144
MAXRECVDATASEGMENTLENGTH: initiator - 131072
target - 65536
LOGIN TIMEOUT : 15
AUTH TIMEOUT : 45
ACTIVE TIMEOUT : 5
IDLE TIMEOUT : 20
PING TIMEOUT : 5
*************************************************************************=
******
I think it's a bit strange that the first target
(iqn.1992-04.com.emc:ax.fcnpr062301759.a0) seems to ignore the idle
timeout setting.
Anyway, I can access data on the SAN without any problems, it's just
annoying with failed/connected messages filling up the log ;-)
Best regards
S=F8ren
|