Gitiles
Code Review Sign In
gerrit.openbmc.org / openbmc / bmcweb / 055713e4b201b049cd6a2d9ca58b89dbf916759f
commit055713e4b201b049cd6a2d9ca58b89dbf916759f[log] [tgz]
authorEd Tanous <etanous@nvidia.com>Wed Jul 17 17:19:36 2024 -0700
committerEd Tanous <ed@tanous.net>Tue Jul 23 20:58:01 2024 +0000
tree4fb959631bcf3f4045cba57dda4654a445ca4f9c
parent84177a2f9562a78bec80d777535457ba44a16ca1 [diff]
Make journal log efficient

Journal logging currently loops over all entries to find even a single
entry.  This was reasonable at the time when bmc couldn't really store a
lot, but now that BMCs are getting significantly more flash storage,
this simplification is insufficient.  In an example system with an
AST2600, this API takes 32 seconds to respond.  This is mediocre for
obvious reasons.

This commit updates to use the sd_journal APIs to let journald do the
skipping, which can use internal details and can be a lot more
efficient.  To get the total size, bmcweb still needs to pull the
sequenceids of HEAD and TAIL to determine the complete size, but this is
still reasonable.

Tested:
Redfish service validator passes.

Various versions of top and skip return the correct result, pulling
various top sizes from 0, omitted to the limit.

https://gerrit.openbmc.org/c/openbmc/openbmc-tools/+/72975

To test all corner cases.

Change-Id: I0456bca4e037529f70eaee0bdd9191e9d5839226
Signed-off-by: Ed Tanous <etanous@nvidia.com>
4 files changed
tree: 4fb959631bcf3f4045cba57dda4654a445ca4f9c
  1. .github/
  2. config/
  3. http/
  4. include/
  5. redfish-core/
  6. scripts/
  7. src/
  8. static/
  9. subprojects/
  10. test/
  11. .clang-format
  12. .clang-tidy
  13. .codespell-ignore
  14. .dockerignore
  15. .eslintignore
  16. .gitignore
  17. .markdownlint.yaml
  18. .openbmc-enforce-gitlint
  19. .prettierignore
  20. .shellcheck
  21. AGGREGATION.md
  22. CLIENTS.md
  23. COMMON_ERRORS.md
  24. DBUS_USAGE.md
  25. DEVELOPING.md
  26. HEADERS.md
  27. LICENSE
  28. meson.build
  29. meson.options
  30. OEM_SCHEMAS.md
  31. OWNERS
  32. README.md
  33. Redfish.md
  34. REDFISH_CHECKLIST.md
  35. run-ci
  36. TESTING.md
README.md

OpenBMC webserver

This component attempts to be a "do everything" embedded webserver for OpenBMC.

Features

The webserver implements a few distinct interfaces:

  • DBus event websocket. Allows registering on changes to specific dbus paths, properties, and will send an event from the websocket if those filters match.
  • OpenBMC DBus REST api. Allows direct, low interference, high fidelity access to dbus and the objects it represents.
  • Serial: A serial websocket for interacting with the host serial console through websockets.
  • Redfish: A protocol compliant, DBus to Redfish translator.
  • KVM: A websocket based implementation of the RFB (VNC) frame buffer protocol intended to mate to webui-vue to provide a complete KVM implementation.

Protocols

bmcweb at a protocol level supports http and https. TLS is supported through OpenSSL.

AuthX

Authentication

Bmcweb supports multiple authentication protocols:

  • Basic authentication per RFC7617
  • Cookie based authentication for authenticating against webui-vue
  • Mutual TLS authentication based on OpenSSL
  • Session authentication through webui-vue
  • XToken based authentication conformant to Redfish DSP0266

Each of these types of authentication is able to be enabled or disabled both via runtime policy changes (through the relevant Redfish APIs) or via configure time options. All authentication mechanisms supporting username/password are routed to libpam, to allow for customization in authentication implementations.

Authorization

All authorization in bmcweb is determined at routing time, and per route, and conform to the Redfish PrivilegeRegistry.

*Note: Non-Redfish functions are mapped to the closest equivalent Redfish privilege level.

Configuration

bmcweb is configured per the meson build files. Available options are documented in meson_options.txt

Compile bmcweb with default options

meson setup builddir
ninja -C builddir

If any of the dependencies are not found on the host system during configuration, meson will automatically download them via its wrap dependencies mentioned in bmcweb/subprojects.

Use of persistent data

bmcweb relies on some on-system data for storage of persistent data that is internal to the process. Details on the exact data stored and when it is read/written can seen from the persistent_data namespace.

TLS certificate generation

When SSL support is enabled and a usable certificate is not found, bmcweb will generate a self-signed a certificate before launching the server. Please see the bmcweb source code for details on the parameters this certificate is built with.

Redfish Aggregation

bmcweb is capable of aggregating resources from satellite BMCs. Refer to AGGREGATION.md for more information on how to enable and use this feature.