Alex Rousskov [Fri, 4 Feb 2011 22:25:45 +0000 (15:25 -0700)]
Quiet down swap out error reporting.
Do not report swap out errors at level 1. When things go wrong, the already
bad situation is made worse by writing lots of error messages to cache.log.
Do not report system error because the errno may be stale or irrelevant.
If error details are needed, the code should save and propagate the actual
errno in addition to the DISK_ERROR or similar status.
When StoreEntry is deleted, we need to release the SwapDir map slot locks it
holds, if any. This is difficult because SwapDir maintains the locks while
Squid Core maintains the entry swap_status. The Core gets swap_status-related
notifications using async calls so it is easy for swap_status to get out of
sync if SwapDir updates the map slot proactively.
The new code no longer releases the slot lock until the associated StoreEntry
is unlinked or gone, even if the slot is known to be unusable and waiting to
be deleted. We also do not rely on swap_status to guess which lock to release;
we use slot state to determine that instead.
Removed rock-specific code from StoreEntry destructor by introducing a general
SwapDir::disconnect(StoreEntry&) API.
Alex Rousskov [Thu, 3 Feb 2011 23:41:32 +0000 (16:41 -0700)]
Revised Slot management in Rock::DirMap.
Old code was occasionally hitting a s.state == Slot::Writing assertion when
closing the writing state. Since I could not find a specific bug that would
lead to this, I decided to simplify state management by moving Slot locking
further away from the Slot state.
Two kinds of Slot locks are now supported: exclusive and shared. These are
implemented using simple atomic counters. To obtain the shared lock, the slot
must also be in a readable, not-marked-for-freeing state (this is where the
lock and the state still overlap). The code should eventually be polished
to use explicit creation-is-acquisition lock objects.
Old code could not cope with Slot deletion event arriving when the Slot was
being written to. We now mark the slot as in need of freeing, regardless of
the slot state. This may need more work to properly cleanup marked slots.
The old code used open/closeForWriting sequences for rebuilding the map from
disk. There were possibly some race conditions in that code. It is now
replaced with an dedicated, simpler, and optimized putAt() method.
Alex Rousskov [Thu, 3 Feb 2011 05:33:05 +0000 (22:33 -0700)]
Support IpcIO timeouts.
Penging IpcIo requests are now stored in two alternating maps: "old" and
"new". Every T seconds, any requests remaining in the "old" map are treated
as timed out. After that check, the current "new" and (now empty) "old" map
pointers are swapped so that the previously "new" requests can now age for T
seconds. New requests are always added to the "new" map. Responses are
always checked against both maps.
This approach gives us access to pending request information and allows to
report errors to the right I/O requestors without creating additional
per-request state attached to a per-request timeout event. The price is (a)
two instead of one map lookups when the response comes and (b) timeout
precision decrease from "about T" to "anywhere from T to 2*T".
Alex Rousskov [Wed, 2 Feb 2011 19:05:25 +0000 (12:05 -0700)]
Fixed Rock MapDir read and write locking:
The IoState object created by openStoreIO() can be used for many reads. Thus,
incrementing read level at open and decrementing it at [each] readCompleted
leads to negative read levels if the stored object need more than one I/O.
Moreover, the only way core Squid can swap in an entry is if an entry has our
fileno set (by our get()). Thus, the slot is already locked for reading by
get(), with the entry responsible for decreasing the read level upon
destruction. We do not need to open/close for reading in
openStoreIO/readComleted.
When writing fails, invalidate the slot before unlocking it.
Alex Rousskov [Wed, 2 Feb 2011 01:49:34 +0000 (18:49 -0700)]
Polished skipping of cache_dirs inactive in a given strand (e.g. Coordinator)
by adding SwapDir::active() method. The directory is active if it makes sense
to call its init/create/get methods in a given strand.
Fixed counting cache_dirs that need dedicated strands. We no longer assume
that all cache_dirs do but use SwapDir::needsDiskStrand() to ask each dir.
The result is stored in Config.cacheSwap.n_strands to optimize NumberOfKids().
Alex Rousskov [Tue, 1 Feb 2011 20:35:42 +0000 (13:35 -0700)]
Call ioCompletedNotification after we are done with the opening sequence,
not in the middle of it. The effect should be the same, but the logs may be
easier to read, and there will be fewer chances of getting into a reentrant
mess of some kind.
Alex Rousskov [Tue, 1 Feb 2011 20:27:13 +0000 (13:27 -0700)]
Do not start rebuilding cache_dir (i.e., loading its index into RAM) until we
complete cache_dir initialization sequence, which ends in not in
Rock::SwapDir::init but in Rock::SwapDir::ioCompletedNotification where we
open the shared map or bail on errors.
It does not make sense to start loading index before the map is configured
because there will be no place to store loaded information.
Alex Rousskov [Tue, 1 Feb 2011 20:18:27 +0000 (13:18 -0700)]
Use Blocking DiskIO module when runnining in a no-daemon mode.
We cannot use IpcIo module in no-daemon mode because there are no diskers
to communicate with. If our implementation is correct, IpcIo module should
contain no shared map or other rock-specific manipulations and, hence,
should not be required for Rock Store to work.
Alex Rousskov [Tue, 1 Feb 2011 08:22:59 +0000 (01:22 -0700)]
Preserve old registration tag when updating registration info.
Sometimes, tagless strand registers self only after its module (like
IpcIoFile) supplies a tag. We need to keep the tag for future tag searches
to succeed.
Alex Rousskov [Tue, 1 Feb 2011 05:01:43 +0000 (22:01 -0700)]
Added IpcIo DiskIO module for communication with remote disk processes via UDS.
Used IpcIo for Rock Store filesystem module.
Added StrandSearch API: Workers use it to ask Coordinator for the right
address (i.e., kid identifier) of the disk process for a given cache_dir path.
If Coordinator does not know the answer, it waits for more disk processes to
register. Implemented using generic tagging of kids (StrandCoord) and
searching for the right tag.
Raised UDS message size maximum to 36K in order to accommodate non-trivial
rock store I/O while we are using UDS messages for I/O content.
Fixed shutdown handling broken by hiding cache_dirs from Coordinator while
switching IamPrimaryProcess() logic to use NumberOfKids() which needs
cache_dir count.
Alex Rousskov [Sun, 30 Jan 2011 23:16:22 +0000 (16:16 -0700)]
Added "disker" processes to be responsible for individual cache_dir I/O.
Determine kid process role based on the process name rather than kid ID.
This allows the process to perform role-specific actions before (or while)
squid.conf is parsed.
Alex Rousskov [Sat, 29 Jan 2011 00:08:52 +0000 (17:08 -0700)]
Added a configuration check to prevent IoState::startWriting() assertions.
Rock::IoState::startWriting() asserts that [padded] write request size does
not exceed the slot size. Padded request size always exceeds the slot size for
slots smaller than the page.
This check may also help avoid using unallocated buffer for padding, but that
part may need more work.
Amos Jeffries [Fri, 14 Jan 2011 14:10:21 +0000 (07:10 -0700)]
Make FTP and CacheMgr obey --disable-auth-basic
When teh proxy has been built with this auth module explicitly disabled
do not add headers indicating that it is available.
The side effect of not having Basic authentication support in the proxy
is that FTP is reduced to depending on URL logins and CacheMgr protected
actions cannot be used.
Amos Jeffries [Fri, 14 Jan 2011 06:15:23 +0000 (23:15 -0700)]
Support configurable status codes for deny_info
This changes the default behaviour of deny_info redirects. Making Squid
automaticaly select 307 or 303 status code where appropriate for HTTP/1.1
clients and 302 for HTTP/1.0 clients or other appropriate cases.
For example;
deny_info 303:http://example.com/ POST
On top of the behaviour change this patch adds capability for admin to
configure deny_info with explicit status codes ranging from 200 to 599.
There are limits placed on the use of each range of status codes:
* 2xx, 4xx and 5xx may only be set when there is a local file or template
being used as body content on the response.
* 3xx status may only be set when there is a URI being used as a redirect
destination.
These limitations are enforced with a configuration hard abort due to:
3xx with a named template and 4xx/5xx with a redirect break with a range
of horrible results to our file loading and output Location: URLs. My
tests ended up with Squid scanning the FS for local files called
http://blah, redirecting the browser to 404:ERR_ACCESS_DENIED, or getting
past those with zero-sized replies and crashes when err is required to
have length.
They are going to take something much more major logic re-plumbing and
maybe deeper cleanup to get the crossover down to safe enough for just a
warning. Given the RFC defined use of each status range I did not think
it worth doing to enable something on the fine edge of non-standard.
Amos Jeffries [Wed, 12 Jan 2011 05:23:00 +0000 (22:23 -0700)]
ftp_eprt directive to disable EPRT extensions in FTP
This allows admin to resolve compatibility problems with old devices which
encounter a range of problems when FTP extensions are used by selectively
disabling any of the extensions individually.
The other EPSV extensions already have enable/disable directives.
Amos Jeffries [Tue, 11 Jan 2011 07:33:27 +0000 (00:33 -0700)]
Bug 2959: remove SAMBAPREFIX dependency
This removes the tricky SAMBAPREFIX variable which passes full-path
information from the squid build machine down to the run-time host
helper.
Such information is not always correct when crossing machines, and the
binaries being run can easily be added to PATH in the run-time host
environment instead.
The net result of doing this is removal of Samba from the build
dependencies and increased availability of the basic_smb_auth and
ext_wbinfo_group_acl helpers.
Amos Jeffries [Sat, 8 Jan 2011 06:23:27 +0000 (19:23 +1300)]
Author: Henrik Nordstrom <hno@squid-cache.org>
Port from 2.7: maximum staleness limits
The default behaviour of Squid is to provide a stale copy (with Warnigng:
header) until an actove response from the origin server causes the object
to be updated or garbage collection causes its removal.
The max_stale direcive and refresh_pattern max-stale=N option allow admin
to set an upper limit on the objects age when serving stale responses.
Amos Jeffries [Mon, 27 Dec 2010 20:25:30 +0000 (13:25 -0700)]
Author: Henrik Nordstrom <hno@squid-cache.org>
Support RFC 5861 Cache-Control: stale-if-error option
The default behaviour for Squid is to present the stale object when
revalidation fails with a 5xx error.
stale-if-error places a maximum limit on how long this stale object may
be sent. After the limit has passed Squid is required to present the 5xx
message to the client.
Original code for Squid-2 was sponsored by Yahoo!.
Original code by Marcello Romani, this version has some additions to
initialize any missing database tables depended on during its startup
phase and some additional polish to fit within the current Squid release.
COPYRIGHT AND LICENSE
Copyright (C) 2008 by Marcello Romani
This library is free software; you can redistribute it and/or modify
it under the same terms as Perl itself, either Perl version 5.8.8 or,
at your option, any later version of Perl 5 you may have available.
Amos Jeffries [Wed, 15 Dec 2010 12:13:01 +0000 (05:13 -0700)]
ext_edirectory_userip_acl: alternative split algorithms
Some compilers do not support dynamically allocated stack space.
Instead perform a scan and hunk copy/wipe of the passed buffers directly.
As a side effect the split is no longer triple-copying data and
double-memset'ing.
Author: Graham Keeling <graham@equiinet.com>
Bug 3113: Squid can eat far too much memory when uploading files
Problem description:
Uploading a large file to a web site on the internet, squid's client
input buffer will increase far faster than it can be emptied to
the target website, and the machine will swiftly run out of memory.
This patch adds the client_request_buffer_max_size configuration
parameter which specifies the maximum buffer size of a client request.