====== QLogic iSCSI HBA and Limitations in Bi-Directional Authentication ======
In the past the QLogic QConvergeConsole (''qaucli'') was used as an administration tool for the hardware initiator part of the QLogic 4000 and QLogic 8200 Series network adapters and iSCSI HBAs. Unfortunately this tool was only supported on the so-called "enterprise Linux distributions" like RHEL and SLES. If you were running any other Linux distribution like e.g. Debian or even one of the BSD distributions you were out of luck.
Thankfully QLogic addressed this support issue indirectly, by first announcing and since then by actually moving from a IOCTL based management method towards the [[http://open-iscsi.org/ | Open-iSCSI]] based management method via the ''iscsiadm'' command. The announcement [[http://www.qlogic.com/Resources/Documents/TechnologyBriefs/Adapters/QLogic_iSCSI_Solution_for_Transitioning_to_the_Open-iSCSI_Model.pdf | QLogic iSCSI Solution for Transitioning to the Open-iSCSI Model]] and the [[http://driverdownloads.qlogic.com/QLogicDriverDownloads_UI/NewSearchResult.aspx?resourceid=MgA4ADMAMQAzAA== | User's Guide IOCTL to Open-iSCSI Interface]] can be found at the QLogic web site.
While trying to test and use the new management method for the hardware initiator via the Open-iSCSI ''iscsiadm'' command, i soon ran into the issue that the packaged version of Open-iSCSI, which is shipped with [[https://packages.debian.org/wheezy/open-iscsi | Debian Wheezy]], is based on the last stable release [[http://www.open-iscsi.org/bits/open-iscsi-2.0-873.tar.gz | v2.0.873]] from Open-iSCSI and is thus hopelessly out of date. The Open-iSCSI package shipped with [[https://packages.debian.org/jessie/open-iscsi | Debian Jessie]] is a bit better, since it's already based on [[https://github.com/mikechristie/open-iscsi/commit/3b4b45001b6d8412aad76a55347de42d30d694f7 | a newer version]] from the projects GitHub repository. Still, the Git commit used there dates back to August 23rd of 2013, which is also fairly old. After updating my system to Debian Jessie, i soon decided to rebuild the Open-iSCSI package from [[https://github.com/mikechristie/open-iscsi/commit/1dfb88a4dfd8e8185491061f557a0c23c522f314 | a much more recent version]] from the projects GitHub repository. With this, the management of the QLogic hardware initiators worked very well via the Open-iSCSI ''iscsiadm'' command and its now enhanced ''host'' mode.
In the ''host'' mode there are now three sub-modes ''chap'', ''flashnode'', ''stats''. See ''man iscsiadm'' and ''/usr/share/doc/open-iscsi/README.gz'' for more details on how to use them. By first calling the host mode without any sub-mode, ''iscsiadm'' prints a list of available iSCSI HBAs along with the host number -- shown in the first pair of square brackets -- associated with each host by the OS kernel:
root@host:~$ iscsiadm -m host
qla4xxx: [1] 10.0.0.5,[84:8f:69:35:fc:70], iqn.2000-04.com.qlogic:isp8214.000e1e37da2c.4
qla4xxx: [2] 10.0.0.6,[84:8f:69:35:fc:71], iqn.2000-04.com.qlogic:isp8214.000e1e37da2d.5
The host number -- in the above example ''1'' and ''2'' -- is used in the following examples showing the three sub-modes:
* The ''stats'' sub-mode displays various statistics values, like e.g. TCP/IP and iSCSI sessions, of the given HBA port:
root@host:~$ iscsiadm -m host -H 1 -C stats
Host Statistics:
mactx_frames: 2351750
mactx_bytes: 233065914
mactx_multicast_frames: 1209409
mactx_broadcast_frames: 0
mactx_pause_frames: 0
mactx_control_frames: 0
mactx_deferral: 0
mactx_excess_deferral: 0
mactx_late_collision: 0
mactx_abort: 0
mactx_single_collision: 0
mactx_multiple_collision: 0
mactx_collision: 0
mactx_frames_dropped: 0
mactx_jumbo_frames: 0
macrx_frames: 4037613
macrx_bytes: 1305799553
macrx_unknown_control_frames: 0
macrx_pause_frames: 0
macrx_control_frames: 0
macrx_dribble: 0
macrx_frame_length_error: 0
macrx_jabber: 0
macrx_carrier_sense_error: 0
macrx_frame_discarded: 0
macrx_frames_dropped: 2409752
mac_crc_error: 0
mac_encoding_error: 0
macrx_length_error_large: 0
macrx_length_error_small: 0
macrx_multicast_frames: 0
macrx_broadcast_frames: 0
iptx_packets: 1694187
iptx_bytes: 112412836
iptx_fragments: 0
iprx_packets: 1446806
iprx_bytes: 721191324
iprx_fragments: 0
ip_datagram_reassembly: 0
ip_invalid_address_error: 0
ip_error_packets: 0
ip_fragrx_overlap: 0
ip_fragrx_outoforder: 0
ip_datagram_reassembly_timeout: 0
ipv6tx_packets: 0
ipv6tx_bytes: 0
ipv6tx_fragments: 0
ipv6rx_packets: 0
ipv6rx_bytes: 0
ipv6rx_fragments: 0
ipv6_datagram_reassembly: 0
ipv6_invalid_address_error: 0
ipv6_error_packets: 0
ipv6_fragrx_overlap: 0
ipv6_fragrx_outoforder: 0
ipv6_datagram_reassembly_timeout: 0
tcptx_segments: 1694187
tcptx_bytes: 69463008
tcprx_segments: 1446806
tcprx_byte: 692255204
tcp_duplicate_ack_retx: 8
tcp_retx_timer_expired: 28
tcprx_duplicate_ack: 0
tcprx_pure_ackr: 0
tcptx_delayed_ack: 247594
tcptx_pure_ack: 247710
tcprx_segment_error: 0
tcprx_segment_outoforder: 0
tcprx_window_probe: 0
tcprx_window_update: 2248673
tcptx_window_probe_persist: 0
ecc_error_correction: 0
iscsi_pdu_tx: 1446486
iscsi_data_bytes_tx: 30308
iscsi_pdu_rx: 1446510
iscsi_data_bytes_rx: 622721801
iscsi_io_completed: 253632
iscsi_unexpected_io_rx: 0
iscsi_format_error: 0
iscsi_hdr_digest_error: 0
iscsi_data_digest_error: 0
iscsi_sequence_error: 0
<
* The ''chap'' sub-mode displays and alters a table containing authentication information. Calling this sub-mode with the ''-o show'' option displays the current contents of the table:
root@host:~$ iscsiadm -m host -H 1 -C chap -o show
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 0
host.auth.username_in =
host.auth.password_in =
# END RECORD
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 1
host.auth.username =
host.auth.password =
# END RECORD
[...]
Why ''show'' isn't the default option in the context of the ''chap'' sub-mode, like it is in many other ''iscsiadm'' modes and sub-modes is something i haven't quite understood yet. Maybe it's a security measure to not accidentially divulge sensitive information, maybe it has just been overlooked by the developers.
Usually, there are already two initial records with the indexes ''0'' and ''1'' present on a HBA. As shown in the example above, each authentication record consists of three parameters. A record index ''host.auth.tbl_idx'' to reference it, a username ''host.auth.username'' or ''host.auth.username_in'' and a password ''host.auth.password'' or ''host.auth.password_in''. Depending on whether the record is used for outgoing authentication of an initiator against a target or the other way around for incoming authentication of a target against an initiator, the parameter pairs ''username''/''password'' **or** ''username_in''/''password_in'' are used. Apparently both types of parameter pairs -- incoming **and** outgoing -- cannot be mixed together in a single record. My guess is that this isn't a limitation in Open-iSCSI, but rather a limitation in the specification and/or of the underlying hardware.
New authentication records can be added with the ''-o new'' option:
root@host:~$ iscsiadm -m host -H 1 -C chap -x 2 -o new
root@host:~$ iscsiadm -m host -H 1 -C chap -o show
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 0
host.auth.username_in =
host.auth.password_in =
# END RECORD
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 1
host.auth.username =
host.auth.password =
# END RECORD
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 2
host.auth.username =
host.auth.password =
# END RECORD
[...]
Parameters of existing authentication records can be set or updated with the ''-o update'' option. The particular record to be set or to be updated is selected in with the ''-x '' option, which references the records ''host.auth.tbl_idx'' value. Multiple parameters can be set or updated with a single ''iscsiadm'' command by calling it with multiple pairs of ''-n '' and ''-v '':
root@host:~$ iscsiadm -m host -H 1 -C chap -x 2 -o update -n host.auth.username -v testuser -n host.auth.password -v testpassword
root@host:~$ iscsiadm -m host -H 1 -C chap -o show
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 0
host.auth.username_in =
host.auth.password_in =
# END RECORD
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 1
host.auth.username =
host.auth.password =
# END RECORD
# BEGIN RECORD 2.0-873
host.auth.tbl_idx = 2
host.auth.username = testuser
host.auth.password = testpassword
# END RECORD
[...]
Finally, existing authentication records can be deleted with the ''-o delete'' option:
root@host:~$ iscsiadm -m host -H 1 -C chap -x 2 -o delete
<
* The ''flashnode'' sub-mode displays and alters a table containing information about the iSCSI targets. Calling this sub-mode without any other options displays an overview of the currently configured flash nodes (i.e. targets) on a particular HBA:
root@host:~$ iscsiadm -m host -H 1 -C flashnode
qla4xxx: [0] 10.0.0.2:3260,0 iqn.2001-05.com.equallogic:0-fe83b6-a35c152cc-c72004e10ff558d4-lun-000002
Similar to the previously mentioned ''host'' mode, each output line of the ''flashnode'' sub-mode contains an index number for each flash node entry (i.e. iSCSI target), which is shown in the first pair of square brackets. With this index number the individual flash node entries are referenced in all further operations.
New flash nodes or target entries can be added with the ''-o new'' option. This operation also needs the information on whether the target addressed via the flash node will be reached via IPv4 or IPv6 addresses. This is accomplished with the ''-A ipv4'' or ''-A ipv6'' option:
root@host:~$ iscsiadm -m host -H 1 -C flashnode -o new -A ipv4
Create new flashnode for host 1.
New flashnode for host 1 added at index 1.
If the operation of adding a new flash node is successful, the index under which the new flash node is addressable is returned.
root@host:~$ iscsiadm -m host -H 1 -C flashnode
qla4xxx: [0] 10.0.0.2:3260,0 iqn.2001-05.com.equallogic:0-fe83b6-a35c152cc-c72004e10ff558d4-lun-000002
qla4xxx: [1] 0.0.0.0:3260,0
Unlike authentication records, the flash node or target records contain a lot more parameters. They can be displayed by selecting a specific record by its index with the ''-x '' option. The ''-o show'' option is the default and is thus optional:
root@host:~$ iscsiadm -m host -H 1 -C flashnode -x 1
# BEGIN RECORD 2.0-873
flashnode.session.auto_snd_tgt_disable = 0
flashnode.session.discovery_session = 0
flashnode.session.portal_type = ipv4
flashnode.session.entry_enable = 0
flashnode.session.immediate_data = 0
flashnode.session.initial_r2t = 0
flashnode.session.data_seq_in_order = 1
flashnode.session.data_pdu_in_order = 1
flashnode.session.chap_auth_en = 1
flashnode.session.discovery_logout_en = 0
flashnode.session.bidi_chap_en = 0
flashnode.session.discovery_auth_optional = 0
flashnode.session.erl = 0
flashnode.session.first_burst_len = 0
flashnode.session.def_time2wait = 0
flashnode.session.def_time2retain = 0
flashnode.session.max_outstanding_r2t = 0
flashnode.session.isid = 000e1e17da2c
flashnode.session.tsid = 0
flashnode.session.max_burst_len = 0
flashnode.session.def_taskmgmt_tmo = 10
flashnode.session.targetalias =
flashnode.session.targetname =
flashnode.session.discovery_parent_idx = 0
flashnode.session.discovery_parent_type = Sendtarget
flashnode.session.tpgt = 0
flashnode.session.chap_out_idx = 2
flashnode.session.chap_in_idx = 65535
flashnode.session.username =
flashnode.session.username_in =
flashnode.session.password =
flashnode.session.password_in =
flashnode.session.is_boot_target = 0
flashnode.conn[0].is_fw_assigned_ipv6 = 0
flashnode.conn[0].header_digest_en = 0
flashnode.conn[0].data_digest_en = 0
flashnode.conn[0].snack_req_en = 0
flashnode.conn[0].tcp_timestamp_stat = 0
flashnode.conn[0].tcp_nagle_disable = 0
flashnode.conn[0].tcp_wsf_disable = 0
flashnode.conn[0].tcp_timer_scale = 0
flashnode.conn[0].tcp_timestamp_en = 0
flashnode.conn[0].fragment_disable = 0
flashnode.conn[0].max_xmit_dlength = 0
flashnode.conn[0].max_recv_dlength = 65536
flashnode.conn[0].keepalive_tmo = 0
flashnode.conn[0].port = 3260
flashnode.conn[0].ipaddress = 0.0.0.0
flashnode.conn[0].redirect_ipaddr = 0.0.0.0
flashnode.conn[0].max_segment_size = 0
flashnode.conn[0].local_port = 0
flashnode.conn[0].ipv4_tos = 0
flashnode.conn[0].ipv6_traffic_class = 0
flashnode.conn[0].ipv6_flow_label = 0
flashnode.conn[0].link_local_ipv6 =
flashnode.conn[0].tcp_xmit_wsf = 0
flashnode.conn[0].tcp_recv_wsf = 0
flashnode.conn[0].statsn = 0
flashnode.conn[0].exp_statsn = 0
# END RECORD
From the various parameters of a flash node or target record, the following are the most relevant in day to day use:
* **flashnode.session.chap_auth_en**: Controls whether the **initiator** should authenticate against the target. This is enabled by default. <
* **flashnode.session.bidi_chap_en**: Controls whether the **target** should also authenticate itself against the initiator. This is disabled by default. <
* **flashnode.session.targetname**: The [[wp>ISCSI#Addressing | IQN]] of the target to be logged into and to be accessed. <
* **flashnode.session.chap_out_idx**: The index number (i.e. the value of the ''host.auth.tbl_idx'' parameter) of the authentication record to be used for authentication of the initiator against the target. <
* **flashnode.conn[0].port**: The TCP port of the target portal. The default is port 3260. <
* **flashnode.conn[0].ipaddress**: The IP address of the target portal. <
The parameter pairs ''flashnode.session.username''/''flashnode.session.password'' and ''flashnode.session.username_in''/''flashnode.session.password_in'' are handled differently than all the other parameters. They are not set or updated directly, but are rather filled in automatically. This is done by setting the respective ''flashnode.session.chap_out_idx'' or ''flashnode.session.chap_in_idx'' parameter to a value which references the index (i.e. the value ''host.auth.tbl_idx'' parameter) of an appropriate authentication record.
Parameters of existing flash nodes or target entries can be set or updated with the ''-o update'' option. The particular record of which the parameters are to be set or to be updated is selected with the ''-x '' option. This references an index number gathered from the list of flash nodes or a index number returned at the time of creation of a particular flash node. Multiple parameters can be set or updated with a single ''iscsiadm'' command by calling it with multiple pairs of ''-n '' and ''-v '':
root@host:~$ iscsiadm -m host -H 1 -C flashnode -x 1 -o update -n flashnode.session.chap_out_idx -v 2 -n flashnode.session.targetname -v iqn.2001-05.com.equallogic:0-fe83b6-d63c152cc-7ce004e1102558d4-lun-000003 -n flashnode.conn[0].ipaddress -v 10.0.0.2
The flash node or target entry updated by this command is shown below in a cut-down fashion for brevity:
root@host:~$ iscsiadm -m host -H 1 -C flashnode -x 1
# BEGIN RECORD 2.0-873
[...]
flashnode.session.chap_auth_en = 1
flashnode.session.discovery_logout_en = 0
flashnode.session.bidi_chap_en = 0
[...]
flashnode.session.targetname = iqn.2001-05.com.equallogic:0-fe83b6-d63c152cc-7ce004e1102558d4-lun-000003
[...]
flashnode.session.chap_out_idx = 4
flashnode.session.chap_in_idx = 65535
flashnode.session.username = testuser
flashnode.session.username_in =
flashnode.session.password = testpassword
flashnode.session.password_in =
[...]
flashnode.conn[0].port = 3260
flashnode.conn[0].ipaddress = 10.0.0.2
flashnode.conn[0].redirect_ipaddr = 0.0.0.0
[...]
# END RECORD
Once configured, ''login'' and ''logout'' actions can be performed on the flash node (i.e. target) with the repective command options:
root@host:~$ iscsiadm -m host -H 1 -C flashnode -x 1 -o login
root@host:~$ iscsiadm -m host -H 1 -C flashnode -x 1 -o logout
Unfortunately the ''iscsiadm'' command will return with a success status on the ''login'' and ''logout'' actions once it has passed them successfully to the HBA. This does not reflect on the status of the actual ''login'' and ''logout'' actions subsequently taken by the HBA against the target configured in the respective flash node! To my knowledge there is currently no information passed back to the command line about the result of the ''login'' and ''logout'' actions at the HBA levels.
Newly established as well as already existing iSCSI sessions via the hardware initiator which were set up with the ''login'' action shown above, are shown along with all other Open-iSCSI session information in the output of the ''iscsiadm'' command in its ''session'' mode:
root@host:~$ iscsiadm -m session
qla4xxx: [1] 10.0.0.2:3260,1 iqn.2001-05.com.equallogic:0-fe83b6-a35c152cc-c72004e10ff558d4-lun-000002 (flash)
qla4xxx: [2] 10.0.0.2:3260,1 iqn.2001-05.com.equallogic:0-fe83b6-a35c152cc-c72004e10ff558d4-lun-000002 (flash)
[...]
The fact that the session information is about a iSCSI session established via a hardware initiator is only signified by the ''flash'' label in the parentheses at the end of each line. In case of a iSCSI session established via a software initiator, the label in the parentheses reads ''non-flash''.
Finally, existing flash nodes can be deleted with the ''-o delete'' option:
root@host:~$ iscsiadm -m host -H 1 -C flashnode -x 1 -o delete
<
The records from both, the ''chap'' and the ''flashnode'' table, are stored in the HBAs flash memory. For the limits on how many entries can be stored in each table, see the specification of the particular HBA.
In my opinion, the integration of management of the QLogic hardware initators into the Open-iSCSI ''iscsiadm'' command improves and simplifies the administration and configuration a lot over the previous IOCTL based management method via the QLogic QConvergeConsole (''qaucli''). It finally opens management access to the QLogic hardware initiators to non-"enterprise Linux distributions" like Debian. Definately a big step in the right direction! The importance of using a current version of Open-iSCSI can -- in my experience -- not be stressed enough. Building and maintaining a package based on a current version from the projects GitHub repository is definitely worth the effort.
One thing i couldn't get to work though was the incoming part of a bi-directional CHAP authentication. In this scenario, not only does the initiator authenticate itself at the target for a iSCSI session to be successfully established, the target also has to authenticate itself against the initiator. My initial thought was, that a setup with bi-directional CHAP authentication should be easily accomplished by just performing the following three steps:
- creating an incoming authentication record with the value of the parameters ''host.auth.username_in'' and ''host.auth.password_in'' set to the respective values configured at the target storage system. <
- setting the value of the flash node parameter ''flashnode.session.bidi_chap_en'' to ''1''. <
- setting the value of the flash node parameter ''flashnode.session.chap_in_idx'' to the value of the parameter ''host.auth.tbl_idx'', gathered from the newly created incoming authentication record in step 1. <
The first two of the above tasks were indeed easily accomplished. The third one seemed easy too, but turned out to be more of a challenge. Setting the flash node parameter ''flashnode.session.chap_in_idx'' to the value of the ''host.auth.tbl_idx'' parameter from the previously created incoming authentication record just didn't work. Any attempt to change the default value ''65535'' failed. Neither was the ''flashnode.session.username_in''/''flashnode.session.password_in'' parameter pair automatically updated with the values from the parameters ''host.auth.username_in'' and ''host.auth.password_in''. Oddly enough bi-directional CHAP authentication worked as long as there only was one storage system with one set of incoming authentication credentials! Adding another set of flash nodes for a second storage system with its own set of incoming authentication credentials would cause the bi-directional CHAP authentication to fail for all targets on this second storage system.
Being unable to debug this weird behaviour any further on my own, i turned to the Open-iSCSI mailing list for help. See [[https://groups.google.com/d/msg/open-iscsi/n1gKU4GogwU/4Qwxg3rQAQAJ | the thread on the Open-iSCSI mailing list]] for the details. Don't be confused by the fact that the thread at the mailing list was initially about getting to work the network communication with the use of jumbo frames. This initial issue was resolved for me by switching to a more recent Open-iSCSI version as already mentioned above. My last question there was not answered publicly on the mailing list, but Adheer Chandravanshi from development at QLogic got in touch via email. Here's his explanation of the observed behaviour:
> [...]
>
> I see that you have added multiple incoming CHAP entries for both hosts 1 and 2.
> But looks like even in case of multiple incoming CHAP entries in flash only the first entry takes effect for that particular host for all
> the target node entries in flash.
> This seems to be a limitation with the flash target node entries.
> So you cannot map different incoming CHAP entry for different target nodes in HBA flash.
>
> In your case, only following incoming CHAP entry will be effective for Host 1 as it's the first incoming chap entry. Same goes for Host
> 2.
>> # BEGIN RECORD 2.0-873
>> host.auth.tbl_idx = 2
>> host.auth.username_in =
>> host.auth.password_in =
>> # END RECORD
>
> Try using only one incoming CHAP entry per host, you can have different outgoing CHAP entries though for each flash target node.
>
> [...]
To sum this up, the QLogic HBAs basically use a first match approach when it comes to the incoming part of a bi-directional CHAP authentication. After finding the first incoming authentication record that is configured, it uses the credentials stored there. Any other -- and possibly more suitable -- records for incoming authentication are ignored. There's also no way to override this behaviour on a case by case basis via the flash node entries (i.e. iSCSI targets).
In my humble opinion this is a rather serious limitation of the security features in the QLogic hardware initiators. No security policy i have ever encountered in any organisation would allow for the reuse of authentication credentials over different systems. Unfortunately i have no further information as to why the implementation turned out this way. Maybe there was no feature request for this yet, maybe it was just an oversight or maybe there is a limitation in the hardware, preventing a more flexible implementation. Unfortunately my reply to the above email with an inquiry whether such a feature would possibly be implemented in future firmware versions has -- up to now -- not been answered.