[201911] Update nat entries to use nat_type to support DNAT Pool changes.

[AkhileshSamineni]

Code changes:

NatMgr:

1. Setting the "APP_NAT_DNAT_POOL_TABLE_NAME" table with destination-ip when first Static NAT/NAPT entry is added at config_db.
2. Incrementing the ref count in "m_natDnatPoolInfo" for same destination-ip for next Static NAT/NAPT entry is added at config_db.
3. Decrementing the reference count in "m_natDnatPoolInfo" for same destination-ip when Static NAT/NAPT entry is deleted at config_db.
4. Deleting the "APP_NAT_DNAT_POOL_TABLE_NAME" table with destination-ip when last Static NAT/NAPT entry is deleted at config_db.
5. It does the same for all IP-Addresses from pool-addresses range in NAT Pool configuration in config_db.

NatOrch:

1. Creating DNAT_POOL entries using nat_type as "SAI_NAT_TYPE_DESTINATION_NAT_POOL" when "APP_NAT_DNAT_POOL_TABLE_NAME" table is created.
2. Deleting DNAT_POOL entries using nat_type as "SAI_NAT_TYPE_DESTINATION_NAT_POOL" when "APP_NAT_DNAT_POOL_TABLE_NAME" table is deleted.
3. Updated code to use nat_type while creating the nat entries.

Addressing the issue - #1234

Depends on :
sonic-net/sonic-utilities#921
sonic-net/sonic-sairedis#616
sonic-net/sonic-swss-common#354

Signed-off-by: Akhilesh Samineni akhilesh.samineni@broadcom.com

[AkhileshSamineni]

Retest this please.

[AkhileshSamineni]

test this please.

[kperumalbfn]

Why do we need DNAT pool entry for the local_ip? I believe DNAT pool lookup is required only for the translated global ip before DNAT/DNAPT table lookup.

[AkhileshSamineni]

User can configure the local IP as the translated source IP in the SNAT direction. Hence in the reverse DNAT direction, we need the DNAT Pool entry for the local IP.

[rlhui]

@kperumalbfn, wondering if this PR looks OK? Thanks.

[AkhileshSamineni]

@rlhui @kperumalbfn Compilation is failing with below error which is unrelated to our PR changes, not sure what could be reason for this failure.

[2020-05-31 02:02:27] [build] fpmsyncd.cpp:3:10: fatal error: logger.h: No such file or directory
[2020-05-31 02:02:27] [build] 3 | #include "logger.h"
[2020-05-31 02:02:27] [build] | ^~~~~~~~~~
[2020-05-31 02:02:27] [build] compilation terminated.

I see some other PR (#1305) also failing with same error.

[kperumalbfn]

@rlhui @kperumalbfn Compilation is failing with below error which is unrelated to our PR changes, not sure what could be reason for this failure.

[2020-05-31 02:02:27] [build] fpmsyncd.cpp:3:10: fatal error: logger.h: No such file or directory
[2020-05-31 02:02:27] [build] 3 | #include "logger.h"
[2020-05-31 02:02:27] [build] | ^~~~~~~~~~
[2020-05-31 02:02:27] [build] compilation terminated.

I see some other PR (#1305) also failing with same error.

@rlhui Any known issue with "logger.h" include file? Can we restart the sanity??

[rlhui]

Retest this please

[arlakshm]

retest this please

[rlhui]

@AkhileshSamineni , please check LGTM failures, thanks.

[kperumalbfn]

@daall LGTM errors on 201911 branch. Is this known one as per your comment on other PR #1355?? If so, can we merge this?

[daall]

@kperumalbfn you are right that it is the same issue, @rlhui you can ignore the LGTM result for this PR as it is not properly set up to build with 201911.