1
'+' = done, '-' = todo, '*' = done except doc
15
stats realm w.ods.org\ statistics
16
stats auth user1:pass1
17
stats auth user2:pass2
18
stats auth user3:pass3
19
stats scope <px_id> | '.'
21
+ allow server-less proxies (for stats)
23
- separate timeout controls
25
+ option 'abortonclose' : if the session is queued or being connecting
26
to the server, and the client sends a shutdown(), then decide to abort
27
the session early because in most situations, this will be caused by
28
a client hitting the 'Stop' button, so there's no reason to overload
29
the servers with unservable requests. However, this is not HTTP compliant
30
and might cause little trouble to some very specific clients used to
31
close immediately after sending the request (no support for KA, which ones?)
33
+ minconn : makes the server's maxconn dynamic, which will be computed as a
34
ratio of the proxy's sessions :
35
srv->effective_maxconn =
36
max(srv->maxconn * px->nbsess / px->maxconn, srv->minconn)
39
+ monitor-uri : specify an URI for which we will always return 'HTTP/1.0 200'
40
and never forward nor log it.
42
+ option ssl-hello-chk : send SSLv3 client hello messages to check the servers
45
- remove unused STATTIME
47
- reference all the include files that must be created, possibly under subdirs :
49
- acl.h => more general ACL work
50
- appcook.h => appsession-related cookies
51
- backend.h => back-end part of the PR_O_* + backend definitions
52
- buffers.h => buffer management relying on memory.h
53
- capture.h => header and cookie capture
54
- cfgparse.h => configuration parser
55
- checks.h => health checks
56
- clireq.h => the client side "request" part of the current sessions.
57
- compat.h => compatibility with other OSes (TCP_NODELAY, ...)
58
- config.h => config parameters, renamed CONFIG_HAP_*, includes defaults.h
59
- controls.h => SN_CACHEABLE, ...
60
- cookies.h => definitions related to cookie management + SN_SCK_*
61
- defaults.h => many default values, might disappear soon after cleanup
62
- frontend.h => front-end part of the PR_O_* + client definitions + listeners
63
- global.h => shared global variables
64
- http.h => HTTP state definitions and transitions
65
- httperr.{hc} => HTTP return codes
66
- libtask.h => task scheduler
67
- libtime.h => time-related definitions
68
- loadbal.h => load balancing algorithms
69
- log.h => log definitions
71
- polling.h => definitions of select(), poll(), INTBITS, ...
72
- queue.h => queue management
73
- regex.h => filtering
74
- servers.h => servers definitions (SRV_*, states, ...)
75
- fd.h => FD_ST* (add FD_DGRAM), RES_*, socket states, etc...
76
- srvreq.h => the server side "request" part of the current sessions.
77
- standard.h => general purpose macros and defines (eg: MIN/MAX, ...)
79
- tuning.h => platform-specific tuning parameters
82
- clarify licence by adding a 'MODULE_LICENCE("GPL")' or something equivalent.
84
- handle half-closed connections better (cli/srv would not distinguish
85
DATA/SHUTR/SHUTW, it would be a session flag which would tell shutr/shutw).
86
Check how it got changed in httpterm.
88
- 3 memory models : failsafe (prealloc), normal (current), optimal (alloc on
91
- wait queues replaced for priority-based trees
93
- ability to assign a prio based on L7 matching
95
- prio-based O(1) scheduler
97
- maxconn reserve for VIP/admins
99
- verify if it would be worth implementing an epoll_ctl_batch() for Linux
101
- balance LC/WLC (patch available)
103
- option minservers XXX : activates some backup servers when active servers
106
- monitor minservers XXX : monitor-net and monitor-uri could report a failure
107
when the number of active servers is below this threshold.
109
- option smtp-chk : use SMTP health checks (avoid logs if possible)
111
- new keyword 'check' : check http xxx, check smtp xxx, check ssl-hello
1
Medium-long term roadmap - 2014/06/19
3
Legend: '+' = done, '-' = todo, '*' = done except doc
6
- POST parameter extraction and size/speed measurement to use in ACLs
8
- return-html code xxx [ file "xxx" | text "xxx" ] if <acl>
10
- return-raw [ file "xxx" | text "xxx" ] if <acl>
12
- avg connect time, response time, connect errors, response errors in stats
14
- add the ability to only dump response errors to more easily detect
15
anomalies without being polluted with attacks in requests.
17
- have multi-criteria analysers which subscribe to req flags, rsp flags, and
18
stream interface changes. This would result in a single analyser to wait
19
for the end of data transfer in HTTP.
21
- implement support for "connection freeze" after accept. A list of frozen
22
connections should be maintained so that it is possible to recycle them
23
when new file descriptors are required.
25
- support for time-ordered priority queues with ability to add an offset
26
based on request matching. Each session will have one ebtree node to be
27
attached to whatever queue the session is waiting in.
29
- add a flag in logs to indicate keep-alive requests ?
31
- make it possible to condition a timeout on an ACL (dynamic timeouts)
33
- forwardfor/originalto except with IPv6
35
- remove lots of remaining Alert() calls or ensure that they forward to
36
send_log() after the fork.
40
- tcp-request session expect-proxy {L4|L5} if ...
42
- http-request track-sc* to avoid having the ugly "if !HTTP" in tcp-request
44
- wait on resource (time, mem, CPU, socket, buffers, server's conn, server's rate, ...)
48
- create internal services and make stats, CLI, etc... part of that.
50
- buddy servers to build defined lists of failovers. Detect loops during
54
server YYY # may replace XXX when XXX fails
56
- spare servers : servers which are used in LB only when a minimum farm
57
weight threshold is not satisfied anymore. Useful for inter-site LB with
58
local pref by default.
60
- add support for event-triggered epoll, and maybe change all events handling
61
to pass through an event cache to handle temporarily disabled events.
63
- evaluate the changes required for multi-process+shared mem or multi-thread
64
+thread-local+fast locking.
66
- ability to decide whether to drain or kill sessions when putting a server
67
to maintenance mode => requires a per-server session list and the change
70
Old, maybe obsolete points :
71
- clarify licence by adding a 'MODULE_LICENCE("GPL")' or something equivalent.
73
- 3 memory models : failsafe (prealloc), normal (current), optimal (alloc on
76
- implement support for event-triggerred epoll()
78
- verify if it would be worth implementing an epoll_ctl_batch() for Linux
80
- option minservers XXX : activates some spare servers when active servers
83
- new keyword 'check' : check http xxx, check smtp xxx, check ssl-hello
85
- initcwnd parameter for bind sockets : needed in kernel first
87
- have a callback function which would be called after a server is selected,
88
for header post-processing. That would be mainly used to remove then add
89
the server's name or cookie in a header so that the server knows it.
92
- outgoing log load-balancing (round-robin or hash among multiple servers)
94
- internal socket for "server XXX frontend:name"
98
- DNS requests on health checks
100
- XML inspection (content-switching for SOAP requests)
102
- sync all stick-tables data, not just serverid.
104
- request, session and user variables
106
- random cookie generator
114
- RAM-based cache for small files
120
- dynamic memory allocation
122
- dynamic weights based on check response headers and traffic response time
124
- various kernel-level acceleration (multi-accept, ssplice, epoll2...)